The 1997 Canadian Election Survey

Technical Documentation

The 1997 Canadian Election Survey

Technical Documentation

David A. Northrup

Institute for Social Research

York University

May 1998

disponible en français

de l'Institut

Conditions of Release

All research based upon these data must include an acknowledgement such as the following:

Data from the 1997 Canadian Election Survey were provided by the Institute for Social Research, York University. The survey was funded by the Social Sciences and Humanities Research Council of Canada (SSHRC), grant number 412-96-0007 and was completed for the 1997 Canadian Election Team of André Blais (Université de Montréal), Elisabeth Gidengil (McGill University), Richard Nadeau (Université de Montréal) and Neil Nevitte (University of Toronto). Neither the Institute for Social Research, the SSHRC, nor the Canadian Election Survey Team are responsible for the analyses and interpretations presented here.

Researchers are requested to forward a copy of any publications or scholarly papers to the Director, Institute for Social Research, York University, 4700 Keele Street, Toronto, Ontario, M3J 1P3 and to André Blais, Départment de Politique Science, Université de Montréal, CP6128 Succ. Centreville, Montréal, H3C 3J7.

Data acquired from the Institute for Social Research may not be re-disseminated outside the recipient institution.

ISR Project Team

Data Collection Jacqueline Davis

Christine Klucha

Mike Scott

Data Set Creation Anne Oram

CATI Development Richard Myles

John Tibert

Project Management David Northrup

Table of Contents

Conditions of Release i

ISR Project Team i

Table of Contents ii

List of Tables iv

1. Study Description 1

1.1 Introduction 1

2. Sample Design 3

2.1 Introduction 3

2.2 Provincial Sample Distribution 3

2.3 Daily Sample Distribution for the Campaign-Period 5

2.4 Selection of Households 8

2.5 Selection of Respondents 9

2.6 Household Weights 9

2.7 Combining Regional and Household Weights for the

Campaign-Period Survey 11

2.8 Post-Election and Mailback Samples 13

3. Data Collection 14

3.1 Introduction 14

3.2 Data Collection Procedures: Campaign-Period and

Post-Election Surveys 14

3.3 Response Rate: Campaign-Period Survey 16

3.4 Re-Interview Rate: Post-Election Survey 19

3.5 Data Collection Procedures: Mailback Survey 20

4. Questionnaire Issues and Data Processing 21

4.1 Introduction 21

4.2 Use of the "RTYPE" and "WAVE" Variables to

Identify Data Sub-Sets 21

4.3 Province Specific Questions 21

4.4 Date Specific Questions 22

4.5 Expectations of the Election Outcome: Identification of

Winning Party and Opposition Party: Explanation of

CATI Code and Errors 22

4.6 Randomization of Question Order 23

4.61 Order Experiments in the Campaign-Period Questionnaire 24

4.62 Question and Response Order Experiments

in the Post-Election Questionnaire 27

4.7 Randomization of Question Wording 29

4.8 Coding of Open-Ended Questions and "Other Specify" Options 31

4.9 Response Time Measurement 36

4.10 Linking Respondents from Three Surveys 36

4.11 Occupational Classification 37

4.12 Listing of Occupations by Occupational Classification Number,

CCDO1980 with accompanying Blishen Socio-Economic

Index Score, 1981 38

4.13 Map of Variables 44

References 56

List of Tables

2.1 Provincial Sample Distribution and Provincial Weights 5

2.2 Completions Per Day: 1997 Campaign-Period Survey 7

2.3 Campaign-Period Survey: Calculation of Household Weights 10

2.4 Explanation of Weights: Campaign-Period Survey 12

3.1 Number of Call Attempts: Campaign-Period and Post-Election Surveys 15

3.2 Final Sample Disposition: 1997 Campaign-Period Survey 17

3.3 Average Percentage of Completions Per Day for Each of the First 27

Days of the Sample Release for the Campaign-Period Survey 18

3.4 Completed Interviews, Response Rates and Re-Interview Rates by

Province: Campaign-Period, Post-Election, and Mailback Surveys 19

1. Study Description

1.1 Introduction

The 1997 Canadian Election Survey (CES) included three survey components.⁽¹⁾ The Campaign-Period Survey (CPS) was completed in the 36 days between the election call, on April 27th, and the last day of the campaign, June 1st. The Post-Election Survey (PES) was completed with respondents to the CPS in the eight weeks after the June 2nd election. The Mailback Survey (MBS) was completed with respondents to the PES from June 19 to October 24, 1997.

Approximately 110 interviews were completed each day of the CPS for a total of 3,949 interviews. Eighty percent, or 3,170, of the CPS respondents completed the PES survey, and 1,857 (59 percent) of the PES respondents completed the mailback survey.

A rolling cross sectional sample release was employed for the campaign-period survey. The sample selection methodology used in the 1997 Canadian Election Survey was similar to that used in the 1988 and 1993 Canadian Election Studies (data collection for these surveys was also completed at ISR). Random digit dialling (RDD) procedures were utilized to select households, and, within households, the birthday selection method was used to select respondents.

Both the English and French interviewing was completed at the Institute's centralized telephone facilities in Toronto using Computer Assisted Telephone Interviewing (CATI) techniques. The Institute uses software from the Computer-Assisted Survey Methods Program (CSM) at the University of California, Berkeley.

In the campaign-period survey respondents were asked about their:

- vote intention and party identification;

- interest in the election, attention paid to polls and the television debates, and what parties/candidates (if any) contacted them during the campaign;

- personal finances and national economic conditions;

- knowledge and rating of the parties and leaders, and how accurately specific terms described party leaders (trustworthy, arrogant, compassionate, in touch with the times, and strong leader);

- personal position, as well as their reading of the main parties' positions, on two key policy issues (cutting taxes or maintaining social programs and how much should be done for Quebec);

- evaluation of how good a job the Liberal government did in a number of policy areas, as well as their review of the performance of the opposition parties;

- expectation about how well each party was doing in the election; and,

- socioeconomic background.

The post-election survey repeated a number of key questions from the CPS, such as leader and party evaluations. Additional topics included government spending (on several social policy areas); attitudes about a number of social issues such as abortion, unions, businesses, education, health care, capital punishment, etc.; a series of questions about support and opposition for Quebec separation, as well as possible outcomes if Quebec did separate; and attitudes towards specific groups in society (big business, unions, feminists, aboriginal peoples, etc.).

The mailback questionnaire dealt primarily with broader political issues and values including questions about respondents' confidence in institutions, the distribution of power between different groups in society, and questions about individual rights and goals of society.

(Copies of the questionnaires are provided under separate cover. Much of the CATI programming language has been omitted, but an explanation of all CATI experiments is included in the questionnaire and in the fourth section of this technical document.)

Details of the sample design, data collection methods, and data set creation are outlined in the remainder of this technical document.

2. Sample Design

2.1 Introduction

The sample for the Canadian Election Survey (CES) was designed to represent the adult population of Canada (Canadian citizens 18 years of age or older who speak one of Canada's official languages, English or French, and reside in private homes⁽²⁾ in the ten Canadian provinces and two territories). Because the mode of data collection for the survey was telephone, the small proportion of households in Canada without telephones were excluded from the sample population.⁽³⁾

2.2 Provincial Sample Distribution and Weights for the Campaign-Period Survey

For purposes of sample design the country was divided into six "regions":

1, the East (Newfoundland, Nova Scotia, Prince Edward Island and New Brunswick);

2, Quebec;

3, Ontario;

4, the Midwest (Manitoba and Saskatchewan);

5, the West (Alberta and British Columbia); and

6, the Territories.

Smaller provinces and the territories were, relative to their population, overrepresented in the sample. The overrepresentation of the smaller provinces facilitates comparison between the six regions. A minimum of 400 interviews were allocated to each region, with a larger allocation of sample going to the larger regions/provinces (Table 2.1). The sample was distributed equally among the provinces when there was more than one province in the region. For example, the 400 cases in the Atlantic region were equally distributed among the provinces of Newfoundland, Prince Edward Island, Nova Scotia and New Brunswick. Unlike Ontario, Quebec was not under-represented in the sample. The larger sample for Quebec (relative to Ontario) ensures enough observations in Quebec so that attitudes towards separation can be factored into analysis.

Because the sample distribution is not proportional to the population of the provinces and territories, the data must be weighted before national estimates are derived. The calculation of the weights to facilitate national estimates is provided in Table 2.1. The weights are calculated by dividing the province or territories' proportion of the households in Canada by the province or territories' proportion of the households in the sample. Ontario has the largest weight (1.5) as the province has 36 percent of Canada's households, but only 24 percent of the sample. In preparing national estimates each Ontario case will count for 1.5 observations in the weighted data set; that is, Ontario is "weighted up" so that the impact of the Ontario sample on national estimates is an accurate reflection of Ontario's proportion of the number of households in Canada. Conversely, for provinces or territories where the weights are very small, for example PEI (.1736) and the Northwest Territories (.0653), the proportion of the sample allocated to the province or territory was greater than that province or territories' proportion of the population. As a result, each case is "weighted down."

Table 2.1. Provincial Sample Distribution and Provincial Weights

	Population^*		Sample
Prov/Ter	HH's (#)	HH's (%)	HH's (#)	HH's (%)	Weight
Nfld.	174,495	1.7	99	2.51	0.6948
PEI	44,478	0.4	101	2.56	0.1736
NS	324,377	3.3	101	2.56	1.2659
NB	253,707	2.5	108	2.73	0.9260
Quebec	2,634,301	26.3	1,034	26.18	1.0040
Ontario	3,638,364	36.3	951	24.08	1.5080
Manitoba	405,120	4.0	203	5.14	0.7866
Sask.	363,149	3.6	211	5.34	0.6784
Alberta	910,391	9.1	481	12.18	0.7461
BC	1,243,894	12.4	473	11.98	1.0366
NWT	16,076	.2	97	.02	.06530
Yukon	10,071	.1	90	.02	.04410
Canada	10,018,423	100.00	3,949	100.00

^* Statistics Canada, 1992. Dwellings and Households: The Nation. Ministry of Industry, Science and Technology, Catalogue No. 93-111, pp 78-89.

Weights that include a correction factor for the unequal probabilities of selection at the provincial and territorial level have been added to the data set to facilitate the production of national estimates. In addition, to facilitate comparisons between Quebec and the rest of Canada, weights have been calculated for Canada without Quebec.

2.3 Daily Sample Distribution for the Campaign-Period Survey

The importance of campaign dynamics in understanding election results has been documented by a number of researchers (Holbrook, 1996; Blais and Boyer, 1996; Johnston, Blais, Gidengil, and Nevitte, 1996; Johnston, Blais, Brady and Crête, 1992; Bartels, 1988; and Brady and Johnston, 1987). By interviewing a cross section of Canadians each day (and including date of interview as a variable in the data set), it is possible to determine the impact of events during a campaign. Using data from the election survey, the analyst can determine if support for specific policy issues, predictions of the results of the election, or ratings of the Prime Minister or the opposition leaders varied, or remained constant, over the course of the election campaign. Similarly, utilization of a rolling cross section sample release facilitates division of the campaign-period data sets into temporal components. For example, analysts can divide the campaign-period data into before and after the leaders' debates.

It is critical to any analysis which includes date of interview as a continuous or contingent variable, that the sociodemographic characteristics of the survey respondents do not systematically vary over time. Because easy-to-reach respondents (people who are more often home and willing to do the interview when first contacted) have different characteristics than hard-to-reach respondents (Groves, 1989; Hawkins, 1975; and Dunkleberg and Day, 1973), it is important that each day of interviewing include a mix of easy and hard-to-reach people.

Assume, for example, that educational achievement is found to covary with attitudes about a specific election issue such as the importance of creating jobs. If more of the interviews at the beginning of data collection were completed with respondents with lower levels of education (and if they were more supportive of job creation efforts as compared to paying down the debt), and if more of the interviews at the end of data collection were completed with respondents with high levels of education (and they were less supportive of job creation efforts), it would be possible to mistake a change in respondent characteristics for a change in attitudes.

Table 2.2 Completions Per Day: 1997 Campaign-Period Survey

Day	Completions	Three Day Average	Five Day Average
April 27	21	--	--
28	85	67	--
29	95	91	80
30	93	97	93
May 1	104	95	93
2	87	92	92
3	86	88	97
4	92	98	106
5	116	118	111
6	147	125	116
7	113	124	118
8	112	109	116
9	103	107	109
10	107	106	117
11	108	124	123
12	157	136	128
13	142	142	132
14	127	132	132
15	126	121	121
16	109	113	115
17	103	108	110
18	112	104	106
19	98	107	111
20	110	113	111
21	132	115	112
22	102	118	112
23	120	106	113
24	96	110	110
25	113	110	114
26	120	118	111
27	122	115	115
28	103	113	115
29	115	111	116
30	114	118	118
31	124	124	--
June 1	135	--	--

The daily variation in the number of completed interviews is expected given the small sample for any one day. However, as seen in Table 2.2, this variation is less pronounced when the number of completed interviews is averaged over a three or five day period.

Every day of sample release was, within provinces and territories, divided into six "sample replicates." Each sample replicate was a random sample of the day's release. Because response to the survey varied by the day of the week (Friday evenings were often least productive while Sunday afternoons were often most productive), and the sample size for any one day was small, there was some modification to the number of replicates released to ensure the number of completions was close to the desired daily goal.

2.4 Selection of Households

A two-stage probability selection process was utilized to select survey respondents. The first stage involved the selection of households by randomly selecting residential telephone numbers. The ideal sampling frame for the campaign-period survey would have been a complete listing of all residential telephone numbers in Canada. Unfortunately, such a listing does not exist and telephone books are not an acceptable surrogate as unlisted numbers (not published in the telephone book by the owner's choice) and numbers for people who have recently moved are not included. Sampling from telephone books would systematically exclude these people from the sample. People who do not have their name in the telephone book are not a random subset of the population (Tremblay, 1982). As a result, ISR employs random digit dialling (RDD) methodology for selecting the telephone numbers.

Use of RDD for selecting telephone numbers gives all households, not just those listed in telephone directories, an equal and known probability of selection. All telephone numbers in Canada consist of an area code, a central office code or exchange (the first three digits of the telephone number), and a suffix or bank (the last four digits of a telephone number). A list of all possible numbers in Canada can be constructed by referring to all telephone books in the country to determine which area code/exchange/bank combinations are in use. For example, once at least one valid telephone number is found in the directory within an area code/exchange/bank combination, e.g., (416) 769-2203, then all numbers from 769-2200 to 769-2299, within the specific area code, are included in the list of all possible telephone numbers. A computer is then used to generate a random sample of telephone numbers from

this listing. As a result, RDD samples include "not-in-service" and "non-residential" telephone numbers. Typically, these non-productive numbers are identified the first time the interviewer calls and most of the interviewer's subsequent efforts are then directed at encouraging respondents to participate in, and then complete, the interview.

2.5 Selection of Respondents

The second stage of the sample selection process was the random selection of a respondent from the selected household. To be eligible for the interview the household member had to be an adult (18 years of age or older) and a Canadian citizen. If there was more than one eligible person in the household, the eligible person who had the next birthday was selected as the survey respondent.⁽⁴⁾ The birthday selection method is used as it ensures a random selection of respondents and it is a much less intrusive way to start an interview than more traditional methods that require a listing of household residents. The less intrusive start makes it easier for the interviewer to secure the respondent's cooperation.

2.6 Household Weights

The probability of an adult member of the household being selected for an interview varies inversely with the number of people living in that household (in a household with only one adult, that adult has a 100 percent chance of selection, in a two adult household each adult has a 50 percent chance of selection, etc.). As a result, it is possible that analyses based on unweighted estimates are biased, as one adult households are over-represented, and larger households are under-represented in the data set. Most practitioners of survey research "weight the data" in order to compensate for the unequal probabilities of selection (one adult households are given a weight of one, two adult households are given a weight of two, three adult households are given a weight of three, etc.).⁽⁵⁾

Conventionally, users of survey data wish to have the same number of observations in the weighted and unweighted data set. This adjustment is made, by determining the number of cases in each household size category that would have been in the sample, if an interview had been completed with each adult member of the household, and then dividing the sample among each household size category according to the proportion of interviews completed in each household size category. The calculation of the household weights for the campaign-period survey is illustrated in Table 2.3

Table 2.3 Campaign-Period Survey: Calculation of Household Weights

HH Size	No. of HH's	Weighted Cases	Adjustment	Weight
1 adult	1,127	1,127	569.78	0.506
2 adults	2,101	4,202	2,124.40	1.011
3 adults	479	1,437	726.50	1.517
4 adults	188	752	380.19	2.022
5 adults	36	180	91.00	2.528
6 adults	15	90	45.50	3.033
7 adults	2	14	7.08	3.539
9 adults^*	1	9	4.55	4.550
Totals	3,949	7,811	3,949.00

^* There were no eight adult households in the sample.

In the campaign-period survey there are 3,949 households in the sample and 1,127 are one-adult households, 2,101 are two-adult households, and 479 are three-adult households, etc. (see variable NADULTS). The weights for each household are calculated as follows. First, the total number of weighted cases is calculated (number of cases times the number of adults in the household). For three-adult households the calculation is: 479 times 3 which gives 1,437 three-adult households in the weighted sample. In the campaign-period survey there are 7,811 weighted cases.

Second, the 7,811 weighted cases are adjusted down to the original sample size of 3,949 (calculated as weighted cases for each household size divided by the weighted sample size times the original sample size). For three-adult households the calculation is: (1,437/7,811) * 3,949 = 726.50.

Third, the weight for each household size is calculated (for each household size, the adjustment to original sample size/number of cases). For three-adult households the calculation is: 726.50/479 = 1.517.

2.7 Combining Regional and Household Weights for the Campaign-Period Survey

Although the weights are provided as part of the data set, users must specify the weights they wish to use in the appropriate programming language before analysing the data. Users are advised to use CPSNWGT1 (campaign-period national weight 1) when national estimates are required, (Table 2.4).

This weight is the product of the household weight and the regional weight. When comparing Quebec to the rest of Canada, the Quebec proportion of the sample should be adjusted using CPSHHWGT (the campaign-period household weight) and CPSNWGT2 (campaign-period national weight 2) should be used for the rest of Canada.⁽⁶⁾ If weights are not invoked the tabulations produced will be for unweighted data.

Table 2.4. Explanation of Weights: Campaign-Period Data Set

	Variable Name	Variable Description	Explanation/When to Use Weight
1	CPSHHWGT	Campaign-Period Household Weight	this weight corrects for unequal probability of selection at the household level
2	CPSPWGT1	Campaign-Period Provincial Weight Number 1	the first provincial weight corrects for unequal probability of selection at the provincial level for the ten provinces and two territories
3	CPSPWGT2	Campaign-Period Provincial Weight Number 2	the second provincial weight corrects for unequal probability of selection at the provincial level after the Province of Quebec has been excluded from the sample
4	CPSNWGT1	Campaign-Period National Weight Number 1	the first national weight combines the household weight and province weight for all ten provinces and two territories
5	CPSNWGT2	Campaign-Period National Weight Number 2	the second national weight combines the household weight and province weight after the Province of Quebec has been excluded from the sample

Separate weights were not prepared for the PES and MBS data sets. The re-interview rates are reasonably high and sample attrition between the surveys was not associated with household size or province and, as a result, it is reasonable to use the CPS weights. Finally, because the weights include fractions that are rounded and missing values vary by item, there will be minor variation in the number of cases for different analytical procedures and subsets of the data. The extent of the rounding problem varies according to the computing program used for analysis. For example, the PC-based version of SPSS has a less effective method for rounding weighted data than the mainframe-based version of the package.

2.8 Post-Election and Mailback Samples

The sample for the post-election survey was comprised of respondents to the CPS. At the end of the CPS, interviewers ensured that they had a first name or some other identifier (such as the respondent's initials or position in the household, e.g., mother). This information, as well as the sex and year of birth of the CPS respondent, and the respondent's telephone number, was recorded on a "cover sheet." At the start of the PES, the cover sheets were put into a random order (shuffled) so that the time of the first call for the PES was not related to the date of interview, or the day of sample release during the CPS.

At the end of the post-election survey, respondents were asked to provide their address so they could be sent the mailback survey. Mailback information was provided by 83 percent of the PES respondents.

3. Data Collection

3.1 Introduction

A description of the data collection procedures is outlined in this section of the technical documentation. Interviewing was completed from ISR's centralized CATI (Computer Assisted Telephone Interviewing) facilities. Each supervisory station is equipped with a video display terminal that reproduces an image of the interviewer's screen and a ROLM CBX telephone communications system. This allows supervisors to monitor (listen to) interviewers' calls and visually verify that the interviewer has recorded the respondent's answers correctly.

3.2 Data Collection Procedures: Campaign-Period and Post-Election Surveys

In order to maximize the chances of getting a completed interview from each sample number, call attempts were made during the day and the evening - for both week and weekend days. Typically, between two and four call attempts were made each day during the first four days that a sample was released. Although over half of the interviews completed in the CPS took three or fewer call attempts, 10 percent of the completed interviews required ten or more calls (Table 3.1). Given the short time that each daily sample was available for calling (10 days), it was important to follow up all possible leads, and as a result, a small number of interviews were completed only after as many as twenty calls were made. The relationship between the number of call attempts and completed interviews in the 1997 election survey parallels that for the 1988 and 1993 election surveys completed at ISR. (The survey data files and accompanying technical documentation for these studies are available at ISR.)

Because the PES did not employ a rolling cross section, and there were no constraints on the number of interviews required per day, it was possible to manage the flow of the sample to interviewers so that most of the calling was completed during the most productive interviewing times. In addition, the respondent knew that an interviewer would be calling back after the election and was expecting the call. As a result we would expect, as was the case in the previous election studies completed at ISR, that the number of call attempts required to complete the interviewing would be less than that required to complete the CPS. However, as detailed below, the re-interview rate for the PES in the 1997 survey was lower than that for the 1993 survey, but the same as that obtained in the 1988 PES (even though the response rate to the 1997 CPS compares favourably to the previous surveys). In an effort to make up for this lower response rate more effort was made to reach respondents at home (and to convert initial refusals). As a result, the number of calls made to obtain a completion, on average, is higher in 1997 than in previous years. In 1993 sixty-two percent of the PES interviews were completed in the first three call attempts, in 1997 only forty-one percent of the interviews were completed on the first three call attempts. Conversely, in 1997 twenty one percent of the PES surveys were completed on the tenth or subsequent call attempt, whereas this percentage was only six percent in 1993. (The variables "CPSATTEM" and "PESATTEM" identify the number of calls required to obtain a completion.)

Table 3.1. Number of Call Attempts: Campaign-Period

and Post-Election Surveys

	CPS		PES
Calls	number	percent	number	percent
1	857	22	510	16
2	780	20	400	13
3	549	14	391	12
4	389	10	348	11
5	318	8	277	9
6	202	5	197	6
7	185	5	150	5
8 - 9	228	6	234	7
10 - 14	273	7	322	10
15 - 37	168	3	313	10
38 - 55			28	1
Totals	3,949	100	3,170	100

Households who refused to participate in the campaign-period survey were contacted a second time and nine percent of the first refusals (179 or 4.5 percent of all CPS interviews) completed the interview on the second or subsequent contact after the initial refusal. (The variables "CPSREFUS and "PESREFUS" identify whether the interview was a "standard" completion or a "converted" refusal.) The limited time that each day's sample was available for calling (as required for the rolling cross section) resulted in a refusal conversion rate considerably lower than the 18 to 23 percent typically achieved in ISR studies. In comparison to the CPS, refusal conversion attempts were almost three times more successful in the PES. While the 118 converted refusals in the PES represent four percent of the PES interviews, they account for twenty-two percent of the initial refusals in the PES survey.

The careful attention to the number and timing of callbacks and refusal conversions is designed to increase the response rate, thereby improving sample representativeness. Many researchers have found that respondents who are "hard-to-reach" and those who "refused" have characteristics that are somewhat different from typical survey responders (Dunkelberg and Day, 1973; Fitzgerald and Fuller, 1982; and McDonald, 1979).

Whether the respondent refused during the initial contact, the number of call attempts, the number of times the telephone was answered and other variables that describe the data collection process are included as part of the data set.

3.3 Response Rate: Campaign-Period Survey

There are numerous ways to calculate response rates in survey research (Groves, 1989; Groves and Lyberg, 1988; Wiseman and Billington, 1984; Frey, 1983; and Dillman, 1978). The method used in this project was conservative; most other ways of calculating the response rate would produce inflated values. The response rate was defined as the number of completed interviews divided by the estimated number of eligible households times 100 percent.

Details on the calculation of the response rate are as follows. Of the 8,748 telephone numbers included in the sample, 6,343 were identified as being eligible households (completions [n=3,949] + refusals [n=2,024] + callbacks [n=370], see Table 3.2). Not eligible households (respondent was unable to speak English or French, was not healthy enough to complete the interview, was not a Canadian citizen, etc. [n=928], and nonresidential and not in service numbers [n=1,071]) accounted for 1,999 of the telephone numbers. It was not possible to determine the eligibility status for 406 of the sample telephone numbers. For response rate calculations, it was assumed that the proportion of these 406 numbers which were eligible household numbers was the same as it was in the rest of the sample.

Table 3.2 Final Sample Disposition: 1997 Campaign-Period Survey

Results	number	percent
completions	3,949	45
refusals	2,024	23
callbacks	370	4
ill/aged/language problem/ absent/not a citizen	928	11
not-in-service & nonresidential	1,071	12
eligibility not determined	406	5
total	8,748	100
participation rate	-	66
completion rate	-	62
household eligibility rate	-	76
estimated number of eligibles	6,651	-
response rate	-	59

This proportion, or "household eligibility rate" was .76 (eligibles [6,343]/(eligibles [6,343] + not eligibles [1,999]) = .76). The estimated total number of eligibles was then computed as 6,651 (6,343 + [.76 x 406] = 6,651). Dividing the number of completions (3,949) by the estimated number of eligibles (6,651) gives a final response rate of 59.4 percent. Many organizations would not include "eligibility not determined" numbers in the denominator for the response rate calculations on the argument that few of these numbers would be eligible households. (See: Groves and Lyberg, 1988 for a debate on this issue.) This version of the response rate, sometimes called a completion rate, calculated as completions/known eligibles is 62 percent (3,949/6,343). Other organizations calculate response rates as the number of completions over the number of completions plus refusals. This version of the response rate, which is sometimes known as the participation rate, is 66 percent (3,949/3,949+2,024).

The response rate for the CPS survey is four to eight percent lower than ISR typically obtains for general population surveys that employ RDD and random selection of the respondent from among all adult household members. The short field period likely accounts for much of the difference as it makes it more difficult to find "hard-to-reach" people at home. The higher than typical percentage of callbacks and eligibility determined telephone numbers in the sample are evidence of the shorter data collection period. In addition, as mentioned above, the shorter data collection period also reduces the effectiveness of "refusal conversion" attempts. In a typical survey, calls are made to households over a two to three week period, while in the CPS the telephone numbers were active for a maximin of ten days.

Table 3.3 Average Percentage of Completions Per Day for Each of the First

27 Days of the Sample Release for the Campaign-Period Survey

# of days	% of comps.	cumulative percent
one	43	43
two	15	58
three	10	68
four	6	74
five	5	79
six	5	84
seven	3	87
eight	5	92
nine	3	95
ten	5	100

Also, the ten day coverage period was truncated for sample released after day 27 of the CPS - the sample released on day 28 was called each day for nine days, the sample released on day 27 for eight days, etc. The percentage of completions obtained, by the number of days the sample was active, for the first 27 days of the CPS is presented in Table 3.3. If we assume the number of completions per day for the last ten days of sample release, would have been about the same as for the first 27 days of calling, the overall response rate would have been three and one-half points higher.

There was variation in the response rate by province and territory. The lowest rates were obtained in Saskatchewan (53 percent) and Quebec (54 percent) and the Territories, Alberta and New Brunswick had the highest rates (between 67 and 72 percent, Table 3.4).

Table 3.4. Completed Interviews, Response Rates, and Re-Interview Rates by Province: Campaign-Period, Post-Election, and Mailback Surveys

	Campaign-Period		Post-Election		Mailback
Prov/Ter	Interviews (#)	Response Rate (%)	Interviews (#)	Re-Interview Rate (%)	Interviews (#)	Re-Interview Rate (%)
Newfoundland	99	66	83	84	48	58
PEI	101	64	86	85	44	51
Nova Scotia	101	67	93	92	64	69
NB	108	68	90	83	58	64
Quebec	1,034	54	801	77	459	57
Ontario	951	59	756	79	453	60
Manitoba	203	62	158	78	89	56
Saskatchewan	211	53	174	82	96	55
Alberta	481	67	407	85	234	57
BC	473	59	371	78	225	61
NWT	97	72	80	82	38	48
Yukon	90	67	71	79	43	61
Canada	3,949	59	3,170	80	1,851	58

3.4 Re-Interview Rate: Post-Election Survey

The post-election re-interview rate is 80 percent. This rate is considerably lower than the 88 percent re-interview rate in the 1993 PES, but it approximates that obtained in the 1988 PES. The response rate for Quebec in the PES, while in the lower range, was within one or two percentage points of four other provinces. The highest reinterview rates for the PES were in the Atlantic region and Alberta and Saskatchewan. The lowest reinterview rates for the mailback were in Prince Edward Island and the Northwest Territories (Table 3.4).

Non-response by CPS respondents to the PES was primarily accounted for by refusals and callbacks. Thirteen percent of all CPS respondents, or just over two thirds of all non-response in the PES, were accounted for by refusals and callbacks. The remaining non-response was accounted for by illness/death of CPS respondents, by never answered telephones, and by changes in telephone numbers (PES respondents had their number changed and the new number was unlisted; the number was changed and the new number listed by the telephone company reached the wrong household; respondent left the household and those remaining in the household either could not or would not provide a new number) or by misdialling in the CPS. (Interviewers are routed, via CATI, to a screen that requests that they verify the telephone number before they proceed to complete the interview; however, given the large volume of calls, some error in dialling is expected and the respondent may not have listened carefully enough to the interviewer when the interviewer asked the respondent if they had correctly dialled the number, e.g., 735-5335 rather than 753-5335).

3.5 Data Collection Procedures: Mailback Survey

At the end of the PES, respondents were asked if they would be willing to provide an address so that a mailback questionnaire could be sent to them. Eighty-three percent of the respondents to the PES provided mailing addresses. All of these 2,627 respondents received the first two mail contacts. The first contact included the questionnaire, a covering letter, and a postage-paid pre-addressed return envelope. The second was a reminder/thank you card (physically like an over-sized post card). The first and second mail contacts were sent from June 19 to August 16 (the mailings were staggered as the first mailbacks were sent prior to the completion of the PES telephone interviewing). Most of the response from these mailings arrived at the Institute within a four week period, at which time a second questionnaire (covering letter and return envelope) was sent only to non-responders. One week later a second reminder card was sent. Finally, during the week of October 21, telephone calls were made to all non-responders. In total, 80 percent of the respondents who provided addresses (or 66 percent of all PES respondents) completed the mailback survey.

4. Questionnaire Issues and Data Processing

4.1 Introduction

This section of the technical documentation provides information about the questionnaire and construction of the data set. A brief description is given of key variables, question order randomization, the coding of open-ended items and the linking of the three data sets. A list of the variables (name and label) is provided. Note that all variables in the Campaign-Period Survey include the prefix "CPS," and the prefixes "PES," and "MBS" are used to indicate that the variable is from the post-election, and mailback survey (respectively).

4.2 Use of the "RTYPE" and "WAVE" Variables to Identify Data Sub-Sets

Questions were survey specific. A frequency tabulation (marginal) for an item from the mail-back survey will include valid cases only for the 1,857 respondents who completed the MBS. A "missing case code" will be assigned to the 2,092 respondents who were part of the Canadian Election Survey but did not complete the MBS. (The 1,857 "valid cases" plus the 2,092 "missing cases" represent the complete sample of 3,949 respondents.) An alternative to including the missing cases is to specify that only a subset of the data is to be used in the analysis. A series of "RTYPE" variables has been created. The variable RTYPE3 for example, identifies respondents to the mailback survey (and RTYPE1 and RTYPE2 identify Campaign-Period and Post-Election survey respondents respectively).

If there was an interest in examining those 1,851 respondents who completed all three surveys, the analyst would select for value 111 of the variable WAVE. A value of 100 in the WAVE variable identifies those 779 respondents who only completed the CPS, and a value of 110 identifies respondents who completed both telephone surveys, but not the mailback.

4.3 Province Specific Questions

A number of survey questions were province specific. For example, when asked to rate leaders or parties in the CPS, respondents from Quebec were not asked about Preston Manning or the Reform Party, and respondents in the remaining provinces and territories were not asked about Gilles Duceppe and the Bloc Québécois. In other circumstances questions were asked only of respondents from Quebec. For example, CPSA2H, rating the importance of "defending the interest of Quebec" was only asked of Quebec respondents. CATI code, such as [if prov ne 14] [goto ch18][ endif] (translated as: if province does not equal Quebec go to the next question), as is found in CPSA2H, identified these province specific questions.

4.4 Date Specific Questions

Questions about the three television debates were asked after May 12 (the English debate starting with item CPSL1), May 13 (the first French debate starting with item CPSL2) and May 18 (the second French debate starting with item CPSL2D). A frequency count for the CPS respondents will produce missing data for all respondents interviewed before these dates.

Two party policy questions are also date specific. The item asking respondents if they knew which party promised to cut unemployment by half by the year 2001 (CPSF14) and the item asking if they knew which party was against recognizing Quebec as a distinct society (CPSF15) were asked in all interviews completed after May 1st (thus a frequency count for the CPS data will identify 500 CPS respondents who did not get asked this question).

The final item added to the CPS survey (CPSJ15) on May 8th, was about the government's decision to have an election at this time (referring to the Manitoba flood).

4.5 Expectations of the Election Outcome: Identification of Winning Party and Opposition Party: Explanation of CATI Code and Errors

During the first stage of data collection there were three CATI problems with the expectations section. Items CPSI2A to CPSI3E were affected. Respondents were to be asked each party's chance of "winning the Election in the whole country" (most seats in Quebec for the Bloc Québécois) in items CPSI2A - CPSI2E. The order the parties were presented to the respondent was randomised. The respondents were asked each party's chances of forming the opposition. The party that the respondent rated as having the best chance of winning in the whole country was to be excluded from the list of parties asked about in the opposition items.

On the 15th of May it was discovered that the plan to exclude the highest ranking party from the opposition items had not been implemented and all parties were being asked the opposition items, regardless of respondent rating. In order to correct this, CATI code was added to the instrument to rank order the parties based on the respondent's answer. This change became effective at the start of interviewing on the 16th of May.

On the 17th of May it became apparent that the rank ordering code was not functioning properly. There were two problems with the code:

One. In certain circumstances respondents were not being asked to rate a party's chances of forming the opposition, even though that party had not been rated the most likely to form the government. A change to the code which corrected the rank ordering problem was made on the 17th of May. This error affected 26 cases completed on the 16th and 17th of May.

Two. When a respondent gave two parties equal chances of forming the government, i.e., a tie, the respondent should have been asked both parties' chances of forming the opposition. This functioned correctly, except that when the tie was created as a result of the respondent asking the interviewer to go back and change a previously given rating, the party with the changed rating was excluded from the opposition items. This problem was not corrected until the 26th of May. The "changed answer tie" affected 44 cases from May 16-25.

In short from April 27 to May 15, the party that the respondent rated as having the best chance of winning in the whole country was not excluded from the list of parties asked about in the opposition items. Between May 16 and May 25, a CATI code error affected 70 cases. Researchers might wish to assign a score of 0 (no chance at all) on the opposition item to the party rated as having the best chances of winning in the whole country.

4.6 Randomization of Question Order

The logical operators resident in CATI were used to randomize the order in which respondents received items in several sections of the questionnaire. Given that order effects have been identified in surveys, but are not always easy to predict (Schuman and Presser, 1981), the order randomization was designed primarily as a precautionary measure to determine what impact, if any, question order had on response.

4.61 Order Experiments in the Campaign-Period Questionnaire

The seven question order experiments in the CPS survey are outlined below. The six question order experiments in the PES and the six question wording experiments in the PES follow.

A: Rotating the Rating of Election Issues

The first question in the CPS was open-ended and asked respondents to identify the most important issue in the election to them PERSONALLY (the emphasis on personal was part of the question). After this question, respondents were asked to rate the importance of eight election issues: national unity (CPSA2A), reducing the deficit (CPSA2B), creating jobs (CPSA2C), cutting taxes (CPSA2D), keeping election promises (CPSA2E), protecting social programs (CPSA2F), fighting crime (CPSA2G), and defending the interests of Quebec (CPSA2H).⁽⁷⁾ The order in which the respondent was asked these eight items was determined by the value of Random Number 1 (CPSRN1). When CPSRN1 was "1", the respondent was asked about national unity first, reducing the deficit second, creating jobs third, etc. When CPSRN1 was "2" the first item in the list was reducing the deficit, the second creating jobs, the third cutting taxes, and the item about nation unity was asked last. As a result, the number of times an item from the list was asked about first, second, third, etc., was approximately the same for each item in the list. A cross tabulation (contingency table) of one of these items by CPSRN1 will allow the analyst to determine the extent to which response varied by question order.

B: Randomization of the Affect of Government Policies on Personal Financial Situation

Respondents were asked if the policies of the federal (CPSC3) or provincial (CPSC4) government made them better off, worse off or if they did not make much difference. Half the respondents were asked about the impact of the federal government's policies on their financial situation first and the provincial government's policies second. This order resulted from CPSRN2 being equal to "1". When CPSRN2 was equal to "2" the order of presentation was reversed so respondents were asked about the affect of the provincial government's policies first and federal government's policies second. Cross tabulations (contingency tables) of CPSRN2 by CPSC3 and CPSRN2 by CPSC4 will allow the analyst to determine what extent, if any, the response to these items varied by question order.

C: Randomization of the Party Leader Ratings

Each respondent was asked to rate four of the five main party leaders (CPSD1A - CPSD1E) on a 0 to 100 scale (Quebec respondents were not asked to rate Preston Manning and respondents in the other nine provinces and two territories were not asked to rate Gilles Duceppe). The order in which the respondent was asked to rate the leaders was determined by CPSRN5. When CPSRN5 had the value "1," respondents (outside of Quebec) were asked to rate the leaders in the following order: Charest, Chrétien, McDonough, Manning. When CPSRN5 had the value of "24" the order of presentation was Manning, McDonough, Chrétien, Campbell. (CPSRN5 included 24 values - "1" to "24" - as there were 24 possible orders.)⁽⁸⁾

Prior to the leader ratings items, respondents were asked if they knew "a lot, a little or nothing at all" about each of the leaders (CPSDR1 - CPSDR5). Respondents who knew nothing at all about a leader were not asked to rate that leader nor their opinion on how well different traits described each leader (see section F below).

D: Randomization of Party Ratings

As was the case for the ratings of party leaders, the 0-100 ratings for parties were randomized (CPSD1G - CPSD1K). Again there were 24 orders and respondents in Quebec were not asked to rate the Reform Party and respondents in the rest of Canada were not asked to rate the Bloc Québécois. The ratings of the parties were controlled by CPSRN6 and the order in which the party ratings were delivered to respondents was independent of the order in which they were asked to rate party leaders.

E: Rotation of Items Rating the Liberals' Performance

A list of seven items (six outside of Quebec), designed to measure the performance of the Liberal government were asked of each respondent (CPSF10A-CPSF10G). These items, with the exception of cutting taxes, are the same issues asked about in CPSA2B-CPSA2H (see part A above). Again the presentation of the issues was rotated. Thus, about equal numbers of respondents were asked the question about "how good a job the Liberals had done in preserving national unity" first, second, third, etc. The rotation of the seven items was controlled by CPSRN3.

F: Randomization of the Leader Traits Battery of Questions

Each respondent was asked how well a set of terms (strong leader, trustworthy, arrogant, compassionate, and in touch with the times) described each party leader (CPH1A - CPSH5E). The order of presentation of the party leaders in this section was randomized using CPSRN7. Again there were 24 orders and the order of presentation of the leaders was independent of the previous ratings questions. Respondents who, in the leader knowledge questions (CPSDR1 - CPSDR5), said they "knew nothing at all about a leader" or who in the leader ratings questions (CPSD1A - CPSD1E) said they "did not know/could not rate/refused to rate a leader" were not asked the traits questions about that leader.

G: Expectation of Vote Outcome in the Riding and in the Country

Respondents were asked what the chances were of each party winning in their riding and the chances of each party winning the country as a whole (CPSI1A - CPSJ2E). When CPSRN4 was "1," respondents were first asked about their riding and second about the country as a whole. The order of presentation was reversed when CPSRN4 was "2." In addition, the order of party presentation was randomized for both the riding and the country questions. For example, when CPSRN8 was "1," (and CPSRN4 was "1") the respondent was asked the chances of the Conservatives winning in their riding, followed by the chances for the Liberals, the NDP, and the Reform (in Quebec, Reform was replaced by Bloc). Conversely when CPSRN8 had a value of 24, the order of parties was reversed (Reform followed by NDP, Liberal, and Conservative). There were 24 different orders for the set of questions about the chances of each party winning in the respondent's riding and 24 orders for the set of questions about the chances of each party winning the country (determined by CPSRN9). Because each respondent was randomly assigned a value for both CPSRN8 and CPSRN9, the order in which

they were asked about the party winning their riding was independent of the order they were asked about a party winning the country as a whole.

4.62 Question and Response Order Experiments in the Post-Election Questionnaire

A: Party Leaders, Parties, and Candidates

The order of presentation of the ratings for the party leaders (PESDR1 - PESDR5), parties (PESC2A - PESC2E), and local candidates (PESC3A - PESC3E) was randomized. As was the case in the CPS, there were 24 unique order presentations for each of these batteries of items (and the order for each battery was independent of the order of the other batteries).

There were important differences between the CPS and PES questionnaires with respect to the party and leader ratings questions. Unlike, as was the case in the CPS, the items measuring knowledge about the party leaders (know a little, a lot, or nothing at all, PESDR1 - PESDR5) were not used as a screen for the leader ratings questions. That is, respondents were asked to rate each leader, even if they answered, in the knowledge questions, that they knew "nothing at all" about a leader. In addition, and also unlike as was the case in the CPS survey, respondents in Quebec were asked how they felt about Preston Manning.⁽⁹⁾ The Manning rating always followed the rating of the other four leaders (which were randomized). In the party ratings questions, respondents in Quebec were asked to rate the Reform party, and respondents in the other Canadian provinces and territories were asked to rate the Bloc Québécois. The Reform Party rating was always asked last in Quebec and the Bloc Québécois rating was always asked last in the other provinces and territories.

B: Spending Cuts

Respondents were asked the extent to which they would cut spending ("a lot, some, or not at all") for seven different areas (defence, welfare, pensions and old age security, health care, unemployment insurance, education, and aid to developing countries, PESE6A - PESE6G). The order in which these areas were presented was rotated, by PESRN9, so that each of the items on the list was asked first-one in seven times, second-one in seven times, etc.

C: Variation in Response Order for the Abortion Item

Respondents were asked which of three positions on abortion (never permitted, permitted after need established, or a woman's personal choice) came closest to their views (PESE5A - PESE5C). Random Number 10 determined in which order these three options were presented to respondents. When PESRN10 was "1" the order was "never, need, choice,", when PESRN10 was "2" the order was "need, choice, never," and when it was "3" the order was "choice, never, need." Frequency counts for the three versions of the questions will allow the analyst to determine the extent (if any) to which the order of the response alternatives affected response.

D: Views on Universality of Government Services

Two positions on universality were offered to respondents (PESE7A and PESE7B). The first suggested government services should not be provided for those who can afford them while the second argued that universality was required to ensure everyone's needs were met. The order in which these two response options were provided to respondents was randomized (PESRN11).

4.7 Randomization of Question Wording

The importance of the way in which issues are framed in question wording has been recognized by survey researchers (Converse and Presser, 1986; and Schuman and Presser, 1981). CATI was also used to vary the wording of several key questions in the Post-Election survey.

A: Identification of Voters

There were two versions of the vote question. The first, PESA2A, was short and direct, "Did you vote in the election" (and was asked when PESRN12 was "1"). The second version of the question (PESA2B) included a more lengthy preamble, which mentioned that "in a democracy" people have "the right to vote" or "not vote" and that "some people end up not voting for one reason or another (asked when PESRN12 was "2"). The extent to which the second version of the question decreased the over-reporting of voting common in surveys (Katosh and Traugott, 1981) can be determined by comparing the frequency distributions for PESA2B and PESA2B.

B: The Temporal Dimension of the Vote Intentions

Respondents were randomly assigned (PESRN2) to either a close-ended (PESA4C) or open- ended (PESA4D) question about when they made up their mind for which party they were going to vote. While the coding scheme for the open-ended version of the question includes response categories from the close-ended questions, additional codes were used to capture the more extensive range of answers provided in the open ended-version of the question. Most common were respondents who answered "I did not make up my mind because always vote for party X," or respondents who said they "made up their mind years before the election was called, etc."

C: Federal/Provincial Government Comparison

There were two versions of the question asking if the respondent's provincial or territorial government, or the federal government, "best looked after the needs of people like you" (PESF14A and PESF14B). The second version of the question (PESF14B) included the response option "or does it not make much difference." The version of the question assigned to respondents was determined by PESRN4.

D: Federal and Provincial Party Identification

The same wording experiment was used for the set of items used to measure federal (PESH1 - PESH9) and provincial (PESH10 - PESH19) party identification. In the federal version of these questions, when PESRN16 was "1", respondents are asked if they "think of themselves as a Liberal, Conservative, NDP or Reform (Bloc in Quebec), or none of these." In the second version of the question (PESH5), respondents are asked if they think of themselves as close to any particular federal party first and then the party they identified with second. The same format is used for the provincial party (PESH10 - PESH19) identification questions.

Respondents who got the first version of the federal party identification question also got the first version of the provincial party identification question.

E: Ranking Goals

Respondents were asked to rank four goals (PESI5A-PESI5F). There were two versions of the list of goals, the first included:

1, maintaining order in the nation;

2, giving people more say in important government decisions;

3, fighting rising prices; and

4, protecting freedom of speech.

All but the third goal were the same on each list. In the second list the third goal was fighting unemployment. Whether the respondent was read the first or second list was determined by Random Number 14 (the first list was read when PESRN14 was "1", the second when it was "2").

4.8 Coding of Open-Ended Questions and "Other Specify" Options

The first question in both the campaign-period (CPSA1) and post-election (PESA1) interviews was open-ended and asked respondents to identify the issue which was most important to them personally in the election. Almost all respondents provided a single response. If a respondent provided more than one response, that could not be coded into a single category, the first response was coded (unless it was not codeable and then the second response was used). The same set of codes (listed below) was used to code both the CPS and PES responses. The list of categories used is extensive and the number of observations in some categories are quite small. However, the use of a large number of categories makes it easier for the analyst to recode the responses into a smaller set of broader categories. An attempt was made, when possible, to use categories developed for the 1993 Canadian Election Study. However, free trade issues, often mentioned in 1993 were infrequent in 1997 whereas there were more mentions of social programs in 1997 than in 1993.

A: Coding Categories for "Most Important Issue" Questions

JOBS AND EMPLOYMENT

10 need/create jobs; reduce unemployment

11 jobs for youth

12 want/need job security (includes things like keeping fisheries open)

13 lack of jobs in resource industry (fishing, farming logging, mining)

14 need more job training, re-training

15 general mention of jobs/unemployment

16 free trade has cost us jobs

FINANCIAL CONCERNS

20 general mention (debt, deficit, etc.)

21 debt - reduce/control/balance

22 debt - eliminate

23 deficit - reduce/control/balance

24 deficit - eliminate

25 transfer payments

26 balance the budget

ECONOMIC CONCERNS

30 general (economy, economic reform)

31 cost of living/inflation, low dollar

32 do something with the interest rate: raise/lower interest rates

33 eco recovery-getting over the recession

34 economic stimulation, initiatives

35 farming/fishing issues (farming, over fishing, costs of transportation)

36 promoting small bus, reduce gov't interference, what will be done for

37 more gov't intervention/fund small bus

38 need to stabilize the economy

HIGH COST OF GOV'T SPENDING

40 general mention high cost of government (too many civil servants)

41 control government spending

42 reduce perks, high salaries, early retire

43 gov't should be accountable for their spending, fiscal responsibility

44 immigration costs/cut back on

45 cut back welfare/clean up abuse

46 helicopter issues

47 early election call/no point to this election/one sided election

48 mention of Manitoba flood

TAXES

50 general mention

51 abolish GST taxes

52 lower GST taxes

53 taxes too high, no new increases

54 give tax break for small business

55 fairer taxation

SOCIAL PROGRAMMES

57 general mention of health care

58 general mention of cutbacks /too many cuts

59 stop health care cuts/more health care $ 60 general mention, keep/protect social

progs

61 old age pensions/security conc' about

cuts

62 child care, increase availability, stop

cuts, more subsidies, keep family benefits

63 social programmes/services, stop cutbacks, more than 1 mention

64 stop social assistance cuts (no

UIC/welfare cuts, reduce waiting time for UIC etc.)

65 education, general mention of concern

66 education, high cost of tuition /stop cuts

67 education - restructure/improve the

system

68 elderly, care of

69 health care concern about availability, afford ability, accessibility

MORAL ISSUES

70 lack of family values, morality

71 abortion issues

72 environmental issues

73 minority issues (equity, aboriginal, gay, women's, human)

74 poverty issues

CRIME AND PUNISHMENT

75 gun control/against gun control- bill C68

76 crime and violence general mentions

77 harsher penalties for criminals, more fairness in justice system

78 young offenders, harsher penalties

79 gun control, no specific mention

UNITY/QUEBEC ISSUES

80 general mention Quebec, bilingualism, sovereignty/future of Quebec/Canada

81 general mention of National Unity

82 maintain National Unity/don't let

Quebec go/Canada should stay as one

country

83 let Quebec go/for separation

84 for sovereignty/independence

85 against sovereignty/independence

86 no distinct society/special status for

Quebec

87 yes distinct society/special status for PQ

88 get rid of Bloc/against Bloc

89 Bloc good for Quebec/want Bloc to win

REPRESENTATION ISSUES

90 get rid of Chrétien/Liberals

91 need more integrity, honesty, accountability from gov't at all levels

92 need stable government one with

foresight strength need good gov't

leaders

93 revamp election process/senate

94 want Chrétien and the liberals to win

95 want a change of gov't

96 representation/recognition for west

97 only concerned with who wins

OTHER

98 don't know, not codeable, other

99 refused B: Coding Categories for PESA8, "What were the Liberals elected to do?"

The set of codes used for the Post-Election question: "What were the Liberals elected to do?"(PESA8) duplicated some of the codes used for the "most important issues questions" used at the start of both the Campaign-Period and Post-Election questionnaires. Again a large number of categories were used and they are organized in such a fashion as to allow for recoding into larger groupings.

JOBS AND EMPLOYMENT

10 create more jobs; reduce

unemployment

11 jobs for youth

12 want/need job security (includes things like keeping fisheries open)

13 lack of jobs in the East

14 need more job training, re-training

15 concentrate on jobs/control

unemployment

FINANCIAL CONCERNS

20 general mention (debt, deficit, etc.)

21 continue reducing debt

22 debt - eliminate

23 continue reducing deficit

24 deficit - eliminate

25 continue their fiscal policies/restraints

26 balance the budget

27 create jobs and reduce deficit/budget (both mentions)

ECONOMIC CONCERNS

30 general mention (economy, economic

reform)

31 fight inflation

32 keep interest rates down

33 improve economy/bring prosperity

34 strengthen/stimulate the economy

35 economy/budget and jobs both mentions

36 need to stabilize the economy

GENERAL COMMENTS

40 general negative comments to lie/screw

us around/spend our money)

41 nothing/not really sure/not much

42 won by default/no choice/best of a

bad lot

43 cater to Quebec

44 cater to Ontario

45 elected by Ont/Quebec (one or both)

46 to try to keep their promises

47 because Chrétien called an early election

48 to beat other parties

49 to change/get new ideas

TAXES

50 general mention

51 abolish GST/taxes

52 cut GST/ taxes

53 keep taxes down

54 jobs and taxes both mentions

55 fairer taxation

SOCIAL PROGS/ SOCIAL SECURITY

57 general mention of health care

58 continue the cutbacks

59 maintain health care /more health care $

60 protect social programmes/social

services

61 protect old age pensions/security

62 protect child care/family benefits

63 jobs and social programmes (both)

64 jobs and health care (both)

65 health care and budget (both)

66 deficit and health care

68 elderly, care of

69 health care concern about availability, afford ability, accessibility

MORAL ISSUES

70 to help Canadians

71 to deal with crime and violence

72 to deal with poverty

UNITY/QUEBEC ISSUES

80 general mention of Quebec,

bilingualism,

sovereignty, independence/future of Quebec/Canada

81 general mention Canadian/National

Unity

82 maintain National/Canadian Unity

83 to stop Quebec Independence

84 unity and economy

85 unity and jobs

86 unity and deficit

87 unity and health care

88 unity and social programmes

REPRESENTATION ISSUES

90 general mention - to carry on

91 honesty, fair & accountable gov't

92 to form a majority gov't

93 to maintain the status quo

94 to continue their mandate

95 to beat the other parties

96 to govern the country

97 to represent the people

OTHER

98 don't know, not codeable, other

99 refused

C: Reasons for Not Voting

Respondents who did not vote in the election were asked if there was a particular reason why they did not do so (PESA3). Five-hundred and twenty-three of the 571 respondents who were asked the question provided an answer that was coded into 13 different categories (including they could not get away from work, had no time, uncertainty with respect to who to vote for, and a sense of cynicism about the whole election process).

D: Respondent's Understanding of What Reform Meant When They Said "All Provinces Should be Treated Equally"

Respondents were asked if they "remembered which party said that all provinces should be treated equally" (PESE29). Those who correctly identified the Reform Party were asked what Reform meant by this (PESE30). The responses were coded into ten categories, the most common four included treat all people equally, all provinces equally, no special status for Quebec, and no distinct status for Quebec.

E: How to Vote on a Day Other than Election Day

Respondents who thought it was possible for someone to vote on a day other than election day (PESA5), were asked how they would do this (PESA5A). Because many respondents gave more than one response their answers were coded into first (PESA5A) second (PESA5A2) and third (PESA5A3) mentioned. Although several categories were used in the coding, most respondents answered that they could vote in an advance poll.

F: Other Specifies

In a number of items, particularly questions about political parties, and in the demographics, interviewers had the option of writing in an "other specify" response. The information provided by interviewers was reviewed and placed into existing categories when appropriate. Observations that remain in the other category in the final data set normally are few in number, or cover such a wide range of possible options that it was not sensible to create specific codes. For a number of the demographic questions, such as ethnicity and language spoken at home, response codes have been added to the data set (note how some response options in the data set are not present in the questionnaire).

In three attitudinal items (what were the Liberals elected to do - CPSF7, what is the best way to fight inflation - CPSF8, the best way to deal with young offenders - CPSJ21, and language usually spoken at home - CPSM14) there were enough other responses to justify the addition of new codes.

4.9 Response Time Measurement

Recent research has explored the relationship between the length of time it takes a respondent to answer a question and how firmly committed they are to their answer (Bassili, 1996; Bassili, 1993; and Bassili and Fletcher, 1991). The questionnaire was programmed, using the clock resident in the CATI system, to measure how long it took respondents to answer a number of questions. The length of time, in hundredths of a second, was stored in a separate variable. Response-time measurement was used for the vote intention question (CPSA4) asked in the CPS (the length of time it took a respondent to answer can be found in variable CPSJF1)

4.10 Linking Respondents from Three Surveys

Considerable effort was made to ensure, within each household, that the same person completed each survey. For example, in the post-election survey, interviewers were provided with the first name, initial, or other identifier (mother, only male in household, etc.) of the respondent who completed the campaign-period survey as well as their sex and year of birth. However, in comparing the name (or identifier), sex, and year of birth for respondents across the surveys, it is possible to isolate cases where there are differences in sex, age, or name (identifier). Each case in the Canadian Election Survey was classified (in the variable RLINK) as being a "goodlink" - including respondents who only completed the CPS - (98 percent), "probable goodlink" (.8 percent), "probable badlink" (0.6 percent), or "Mailback badlink" (0.4 percent). The following conventions were used in the classification.

i. When the name (or identifier), age, and sex were the same in all five surveys the case was classified as a "goodlink."

ii. When the name was different, or there was a change in sex, the case was coded as a "probable badlink."

iii. When the age was different the case was coded as a "probable badlink", with the exception noted in point iv.

iv. When age was different but there was the possibility of an interviewer entry error (for example, year of birth was recorded as 1945 in the first survey and 1954 in the second survey) and there was strong supporting evidence that the same person was interviewed (for example, there was only one male adult in the household who had the correct name), the case was classified as a "probable goodlink".

v. When the linking problems were specific to the mailback survey, the case was classified as a "Mailback badlink."

Analysts who are working with the data may wish to consider dropping the "probable badlink" cases from the data set.

4.11 Occupational Classification

Respondents, in the CPS, who were currently working (including self-employed), laid-off, or unemployed (CPSM4) were asked their current or last occupation (CPSM6). The description of their occupation, recorded as open-ended text by the interviewer, was coded into a 4-digit occupation category using Statistics Canada's "Standard Occupational Classification, 1980." For example, respondents who described their occupation as a high school teacher were assigned a code of 2733. Those who described their occupation as a homemaker were assigned a value of 9994; those who described their occupation as being a student were assigned 9995, disabled a 9996, retired a 9997, don't know a 9998 and if the respondent refused to answer, or provided an answer that was not codeable, the variable was assigned a 9999.

The codebook for the 1980 occupation classifications is contained in this section. Appended to each occupation is a socio-economic index score. These indices are commonly referred to as "Blishen Scores" and are based on the male labour force population who reported an occupation in the 1981 Canadian Census. The development of the scale is reported in Blishen, Carroll and Moore (1987).

Another well-known socio-economic index was developed by Pineo, Porter and McRoberts (1977), based on the 1971 Canadian Census. This index was updated in 1985 to reflect the 1981 Census and is reported in McMaster University (1985).

The data file contains two socio-economic indices. The Blishen Scores are contained in the variable "BLISH81R" and are identical to those shown in the detailed codebook. The Pineo/Porter/McRoberts scores are contained in the variable "PINPORR". The full set of

SPSS recode statements used to create these two indices is available from the Institute on request.

4.12 Listing of Occupations by Occupational Classification Number, CCDO

1980 with accompanying Blishen Socio-Economic Index Score, 1981

CCDO Blishen CCDO Blishen

Number Description Score Number Description Score

1111 Members of legislative bodies 55.08

1113 Government administrators 66.84

1115 Post office management 38.19

1116 Inspectors+regulatory officers, gov't 56.42

1119 Officials,admin. unique to gov't:n.e.c. 59.94

1130 General managers,other senior officials 71.62

1131 Mgmt:natural sciences and engineering 79.23

1132 Mgmt:social sciences+related fields 62.53

1133 Adminis. in teaching, related fields 78.34

1134 Adminis. in medicine and health 68.89

1135 Financial management 60.65

1136 Personnel, industrial relations mgmt 62.87

1137 Sales and advertising management 50.07

1141 Purchasing management 50.83

1142 Services management 40.99

1143 Production management 57.57

1145 Management:construction operations 55.91

1146 Farm management 32.06

1147 Management:transport and commun-

ications operations 61.01

1151 Other management:mines+oil wells 66.39

1152 Other mgmt:durable goods manuf. 56.56

1153 Other mgmt:non-durable goods manuf. 54.91

1154 Other management:construction 49.40

1155 Oth. mgmt:transp.+commun. 56.38

1156 Other management:trade 47.79

1157 Other management:service 52.49

1158 Other mgmt:other industries 56.83

1171 Accountants, auditors and other

financial officers 59.44

1173 Organization and methods analysts 65.98

1174 Personnel and related officers 57.19

1175 Purchasing officers+buyers,except

wholesale+retail trade 52.23

1176 Inspectors+regulatory officers:n.e.c. 52.51

1179 Related to mgmt and admin:n.e.c. 57.55

2111 Chemists 63.47

2112 Geologists 71.01

2113 Physicists 73.00

2114 Meteorologists 70.66

2117 Physical sci.:technologists+technicians 54.05

2119 Physical sciences:n.e.c. 41.81

2131 Agriculturists and related scientists 62.19

2133 Biologists and related scientists 65.63

2135 Life sciences:technologists+technicians 52.86

2139 Life sciences:n.e.c. 51.01

2141 Architects 68.12

2142 Chemical engineers 72.47

2143 Civil engineers 71.70

2144 Electrical engineers 70.48

2145 Industrial engineers 64.07

2146 Agricultural engineers 64.22

2147 Mechanical engineers 68.37

2151 Metallurgical engineers 71.05

2153 Mining engineers 72.80

2154 Petroleum engineers 74.67

2155 Aerospace engineers 65.79

2156 Nuclear engineers 75.44

2157 Community planners 65.11

2159 Professional engineers:n.e.c. 70.27

2160 Supervis.:oth. occup.in architec.+ engin 62.97

2161 Surveyors 46.22

2163 Draughting 53.83

2164 Architectural technolog.+technic. 55.82

2165 Engineering technologists+technicians 56.57

2169 Oth. occup. in architec.+engineer.:n.e.c. 35.47

2181 Math.,statisticians+actuaries 61.91

2183 Systems analysts,computer prog.., rel. 60.73

2189 Math.,stat.,systems analysis, rel.:n.e.c. 48.24

2311 Economists 69.18

2313 Socio.,anthropologists+rel. social sci. 63.09

2315 Psychologists 65.36

2319 Social sciences:n.e.c. 49.87

2331 Social workers 60.11

2333 Welfare and community services 36.89

2339 Social work and related fields:n.e.c. 44.39

2341 Judges and magistrates 93.27

2343 Lawyers and notaries 75.60

2349 In law and jurisprudence:n.e.c. 48.72

2350 Superv.:library,museum+archival sci. 57.97

2351 Librarians,archivists+conservators 55.40

2353 Techn. in library,museum+archival sci. 51.11

2359 Library,museum+archival sci.:n.e.c. 37.70

2391 Educational+vocational counsellors 67.61

2399 Other social sci.+rel. fields:n.e.c. 51.54

2511 Ministers of religion 52.84

2513 Nuns and brothers 42.17

2519 Religion:n.e.c. 43.27

2711 University teachers 75.87

2719 University teaching+related:n.e.c. 46.83

2731 Elementary+kindergarten teachers 63.64

2733 Secondary school teachers 70.19

2739 Elemen./secon. teach.+rel.:n.e.c. 43.38

2791 Comm. college+vocat. school teach. 66.03

2792 Fine arts school teachers:n.e.c. 40.93

2793 Post-secondary school teachers:n.e.c. 67.05

2795 Teachers of exceptional students:n.e.c. 58.09

2797 Instructors and training officers:n.e.c. 49.94

2799 Other teaching and related:n.e.c. 53.23

3111 Physicians and surgeon 101.32

3113 Dentists 101.74

3115 Veterinarians 72.24

3117 Osteopaths and chiropractors 70.24

3119 Health diagnosing and treating:n.e.c. 57.21

3130 Supervisors:nursing,therapy+rel.assis. 63.51

3131 Nurses,regist.,grad.+nurses-in-train. 55.26

3132 Orderlies 38.68

3134 Registered nursing assistants 46.51

3135 Nursing attendants 33.60

3136 Audio and speech therapists 62.36

3137 Physiotherapists 56.56

3138 Occupational therapists 55.23

3139 Nursing,therapy+rel. assisting:n.e.c. 40.44

3151 Pharmacists 64.39

3152 Dietitians and nutritionists 59.31

3153 Optometrists 79.63

3154 Dispensing opticians 48.55

3155 Radiolog. technologists+technicians 56.78

3156 Med lab. technologists+technicians 55.79

3157 Denturists 59.02

3158 Dental hygienists+dental assistants 45.02

3161 Dental laboratory technicians 45.15

3162 Respiratory technicians 59.05

3169 Other in medicine and health:n.e.c. 39.86

3311 Painters,sculptors and related artists 36.88

3313 Product and interior designers 43.47

3314 Advertising and illustrating artists 47.23

3315 Photographers and cameramen 44.66

3319 Fine+com. art,phot.+rel. fields:n.e.c. 40.57

3330 Prod.+direct.,perf.+audio-vis. arts 57.04

3331 Conductors,composers+arrangers 42.01

3332 Musicians and singers 36.58

3333 Music+musical entertain. rel.:n.e.c. 32.35

3334 Dancers and choreographers 32.94

3335 Actors/actresses 42.94

3337 Radio and television announcers 46.43

3339 Performing and audio-visual arts:n.e.c. 37.54

3351 Writers and editors 54.58

3355 Translators and interpreters 57.30

3359 Writing:n.e.c. 50.15

3360 Supervisors:sports and recreation 38.48

3370 Coach.,train.,instr.+manag.:sport+rec. 36.71

3371 Referees and related officials 23.77

3373 Athletes 40.36

3375 Attendants:sport and recreation 24.93

3379 Sport and recreation:n.e.c. 25.74

4110 Supervisors:stenographic and typing 46.00

4111 Secretaries and stenographers 41.82

4113 Typists and clerk-typists 38.47

4130 Supervis.:bookkeep.,account-rec.+rel. 45.39

4131 Bookkeepers and accounting clerks 40.28

4133 Cashiers and tellers 28.31

4135 Insurance,bank and other finance clerks 40.51

4137 Statistical clerks 41.79

4139 Bookkeep.,account-record.+rel.:n.e.c. 40.23

4140 Supervis.:office mach.+e.d.p.equ.oper. 51.16

4141 Office machine operators 37.39

4143 Electronic data-processing equip. oper. 41.93

4150 Supervisors:mat. record.,sched.+dist. 44.50

4151 Production clerks 43.11

4153 Shipping and receiving clerks 34.11

4155 Stock clerks and related 35.46

4157 Weighers 32.07

4159 Mater. recording,sched.,distrib.:n.e.c. 31.89

4160 Superv.:library,file+corres. clerks+rel. 50.57

4161 Library and file clerks 34.85

4169 Library,file and corres.clerks+rel.:n.e.c. 43.50

4170 Superv.:recep.,info.,mail+message dist. 46.46

4171 Receptionists and information clerks 35.04

4172 Mail carriers 42.29

4173 Mail and postal clerks 38.15

4175 Telephone operators 33.25

4177 Messengers 28.82

4179 Recep.,info.,mail+mes. distrib.:n.e.c. 34.90

4190 Supervis.:other clerical+related:n.e.c. 47.88

4191 Collectors 43.10

4192 Claim adjusters 41.70

4193 Travel clerks,ticket,station, freight agen. 44.92

4194 Hotel clerks 31.63

4195 Personnel clerks 45.22

4197 General office clerks 37.93

4199 Other clerical and related:n.e.c. 39.01

5130 Supervisors:sales:commodities 41.01

5131 Technical sales and related advisers 57.89

5133 Commercial travellers 50.52

5135 Sales clerks, salesp.:commod.:n.e.c. 30.93

5141 Street vendors+door-to-door sales 29.95

5143 Newspaper carriers and vendors 17.81

5145 Service station attendants 21.47

5149 Sales:commodities:n.e.c. 29.16

5170 Supervisors:sales:services 56.44

5171 Insurance sales 50.18

5172 Real estate sales 49.99

5173 Sales agents+ traders:securities 58.62

5174 Advertising sales 47.26

5177 Business services sales 52.09

5179 Sales:services:n.e.c. 44.56

5190 Supervisors:other sales 44.32

5191 Buyers,wholesale and retail trade 46.08

5193 Route drivers 35.73

5199 Other sales:n.e.c. 32.84

6111 Fire-fighting 51.17

6112 Police officers+detectives,gov't 58.78

6113 Police agents+investigators,private 46.60

6115 Guards and related security 31.95

6116 Commissioned officers,armed forces 62.19

6117 Other ranks,armed forces 41.69

6119 Protection service:n.e.c. 33.20

6120 Supervis.:food+bev. prep.+rel. serv. 34.64

6121 Chefs and cooks 25.56

6123 Bartenders 29.24

6125 Food and beverage serving 23.31

6129 Food and bev. prep.+ rel. serv.:n.e.c. 26.52

6130 Supervis.:in lodging+oth. accom. 31.36

6133 Lodg. cleaners,except priv. househo. 21.37

6135 Sleeping-car and baggage porters 27.46

6139 Lodging and other accom.:n.e.c. 26.13

6141 Funeral directors,embalmers+ rel. 47.32

6142 Housekeepers,servants and related 22.08

6143 Barbers,hairdressers and related 35.62

6144 Guides 32.87

6145 Travel+rel. attend.,exc. food+bev. 48.83

6147 Child-care occupations 23.70

6149 Personal service:n.e.c. 25.53

6160 Supervis.:apparel+furnishings ser. 34.28

6162 Laundering and dry cleaning 25.90

6165 Pressing 24.49

6169 Apparel+furnishings service:n.e.c. 24.49

6190 Supervisors:other service 37.46

6191 Janitors, charworkers and cleaners 26.36

6193 Elevator-operating 32.21

6198 Labouring+oth. elemental:oth. serv. 21.24

6199 Other service:n.e.c. 27.60

7113 Livestock farmers 29.59

7115 Crop farmers 31.32

7119 Farmers:n.e.c. 27.92

7180 Fore./w:oth. farm.,hort.+ anim. husb. 38.95

7183 Livestock farm workers 25.36

7185 Crop farm workers 22.04

7195 Nursery and related workers 26.99

7196 I.t.g.+s.:other farm.,horticul.+anim. husb 25.71

7197 Farm machinery operators 23.76

7199 Other farming,horti.+animal husb.n.e.c. 23.34

7311 Captains+other officers:fishing vessels 36.35

7313 Net,trap and line fishing 24.59

7315 Trapping and related 19.02

7319 Fishing,trapping and related:n.e.c. 22.73

7510 Foremen/women:forestry and logging 45.16

7511 Forestry conservationist 34.14

7513 Timber cutting and related 25.23

7516 Log inspecting,grading,scaling+rel. 44.19

7517 Log hoisting,sorting,moving+ rel. 34.57

7518 Labour.+oth. elemental:forestry, log. 25.34

7519 Forestry and logging:n.e.c. 32.30

7710 Forem/w:min.+quar. incl.oil+gas field 54.07

7711 Rotary well-drilling and related 42.43

7713 Rock and soil-drilling 40.23

7715 Blasting 40.43

7717 Min.+quarry.:cut.,handl.+loading 39.56

7718 Lab.+oth. elem. min + quarry incl. oil+gas 34.73

7719 Min.&quarry. incl. oil&gas field:n.e.c. 40.74

8110 Foremen/women:mineral ore treating 51.56

8111 Crushing and grinding:mineral ores 39.45

8113 Mix.,separat.,filter.&rel.:mineral ores 42.59

8115 Melting and roasting:mineral ores 43.35

8116 I.t.g.+s.:mineral ore treating 45.92

8118 Labour.+oth. element.:miner. ore treat. 37.94

8119 Mineral ore treating:n.e.c. 40.81

8130 Foremen/women:metal processing+rel. 51.27

8131 Metal smelting,converting and refining 40.30

8133 Metal heat-treating 39.33

8135 Metal rolling 41.18

8137 Moulding,coremaking and metal casting 36.45

8141 Metal extruding and drawing 36.41

8143 Plating,metal spraying and related 33.89

8146 I.t.g.+s.:metal processing 44.50

8148 Labouring&other elemental:metal proc. 36.06

8149 Metal processing and related:n.e.c. 38.29

8150 Forem./w:clay,glass+stone pro.,for.+rel 44.48

8151 Furnacemen,kiln work.:clay,glass,stone 36.43

8153 Separ.,grind.,crush.,mix.:clay,glass,stone 34.81

8155 Forming:clay,glass and stone 34.85

8156 I.t.g.+s.:clay,glass+stone process.+form 37.98

8158 Labour.+oth. elem.:clay,glass+stone

process.+form. 31.45

8159 Clay,glass+stone proc.,form.+rel.:n.e.c. 36.07

8160 Forem./w:chem.,petrol,rubb., plast. +rel.mat.proc. 49.77

8161 Mixing,blending:chemicals&rel. mat. 36.19

8163 Filter.,strain.+separat.:chem.+rel.mat. 40.14

8165 Distill.,subl.+carbon.:chem.+rel.mat. 51.21

8167 Roasting,cook.,dry.:chem.+rel.mat. 39.76

8171 Crushing,grinding:chem.+rel.mat. 34.69

8173 Coating,calendering:chem.rel.mat. 32.40

8176 I.t.g.+s:chem.,petrol.rubber,plast.+

el.mat.process. 43.64

8178 Labour.+oth.elem.:chemicals,petr.rub.

plas.+rel.mat.proc. 32.50

8179 Chem.,petrol.,rubber,plast.+rel.mat.

process.n.e.c. 40.75

8210 Foremen/w:food,bev.+rel. processing 41.92

8211 Flour and grain milling 34.77

8213 Baking,confectionery making and rel. 30.55

8215 Slaughtering,meat cut.,can.,cur.+pack. 33.82

8217 Fish canning,curing and packing 20.38

8221 Fruit+veg. canning,preserv.+pack. 23.18

8223 Milk processing and rel. occup. 37.03

8225 Sugar processing and rel. 36.76

8226 I.t.g.+s.:food,beverage+rel. process. 34.09

8227 Beverage processing and related 40.13

8228 Lab.+oth. elem.:food,bev.+rel. proc. 24.92

8229 Food,beverage and rel. proc.:n.e.c. 32.32

8230 Forem./w:wood proc.,exc. pulp+paper 44.20

8231 Sawmill sawyers and related 33.71

8233 Plywood making and related 34.66

8235 Wood treating 35.92

8236 I.t.g.+s.:wood proc.,exc. pulp+paper 38.91

8238 Labour.+oth. elem.:wood proc.,except

pulp+paper 29.71

8239 Wood process.,exc. pulp+paper:n.e.c. 34.87

8250 Foremen/women:pulp+paper+rel. 52.46

8251 Cellulose pulp preparing 44.18

8253 Papermaking and finishing 43.92

8256 I.t.g.+s.:pulp and papermaking 46.10

8258 Labour.+oth. elem. work:pulp+paper 39.32

8259 Pulp+papermaking and related:n.e.c. 39.74

8260 Foremen/women:textile processing 40.71

8261 Textile fibre preparing 29.13

8263 Textile spinning and twisting 28.74

8265 Textile winding and reeling 27.90

8267 Textile weaving 30.36

8271 Knitting 27.82

8273 Textile bleaching and dying 32.29

8275 Textile finishing and calendering 29.16

8276 I.t.g.+s.:textile processing 30.21

8278 Labour+oth. elemental:textile proc. 27.40

8279 Textile processing:n.e.c. 29.65

8290 Foremen/women:other processing 43.35

8293 Tobacco processing 36.65

8295 Hide and pelt processing 28.42

8296 I.t.g.+s.:other processing 35.64

8298 Labouring+other elemental:other proc. 28.78

8299 Other processing:n.e.c. 38.18

8310 Foremen/women:metal machining 50.89

8311 Tool and die making operations 48.15

8313 Machinist and machine tool setting-up 43.99

8315 Machine tool operating 38.43

8316 I.t.g.+s.:metal machining 42.47

8319 Metal machining:n.e.c. 36.62

8330 Forem./w:metal shap.,form.,exc. machin 49.19

8331 Forging 37.68

8333 Sheet metal workers 40.36

8334 Metalworking-machine operators:n.e.c. 34.06

8335 Welding and flame cutting 41.42

8336 I.t.g.+s.:metal shap.,form.,exc. machining 43.19

8337 Boilermakers,platers+struct metal work 43.58

8339 Metal shap.+form.,except mach.:n.e.c. 34.61

8350 Foremen/women:wood machining 41.47

8351 Wood patternmaking 42.52

8353 Wood sawing and related:n.e.c. 30.68

8355 Planing,turning,shaping+rel wood mach 31.62

8356 I.t.g.+s.:wood machining 34.03

8357 Wood sanding 27.51

8359 Wood machining:n.e.c. 31.82

8370 Forem./w.:clay, glass, sto.+rel.mat. mach. 43.15

8371 Cutting+shap.:clay,glass,stone+rel. mat 33.26

8373 Abra.+pol.:clay, glass, sto.+rel. mat.:n.e.c. 32.88

8376 I.t.g.+s.:clay,glass,stone+rel. mat.mach. 36.21

8379 Clay,glass,stone+rel.mat. mach.:n.e.c. 35.01

8390 Foremen/women:other mach+rel.:n.e.c. 46.88

8391 Engravers,etchers and rel.:n.e.c. 32.27

8393 Filing,grind.,buff.,clean.+polish.:n.e.c. 35.40

8395 Patternmakers and mouldmakers:n.e.c. 42.82

8396 I.t.g.+s.:other machining and related 33.55

8399 Other machining and related:n.e.c. 32.48

8510 Forem./w:fabr.+ assam.:metal prod.n.e.c. 49.97

8511 Engine+rel.equip. fab.+assam.:n.e.c. 36.00

8513 Motor vehicle fabricating+assam:n.e.c. 36.86

8515 Aircraft fabricating+assembling:n.e.c. 43.57

8523 Ind.,farm,const.+oth.make.equip.+mach.: fab.+assam:n.e.c. 36.35

8525 Bus.+ comm. mach. :fabric.+ assam. n.e.c. 35.56

8526 I.t.g.+s.:fabric.+assam.metal prod.n.e.c. 43.88

8527 Prec. instr.+rel.equip:fab.+assam.n.e.c. 36.24

8528 Lab.+oth.el.fabri+assam.met. prodn.e.c. 31.03

8529 Other fabric.+assam.:metal prod.:n.e.c. 33.83

8530 Fore./w.:fab.,ass.,inst.+rel.ele.+rel.eg. 50.36

8531 Elect.+rel. equip.:fabric.+assembl. 33.31

8533 Elect.+rel. equip.:insta.+repair.:n.e.c. 48.14

8534 Electronic+rel. equip.:fabric.+assam. 32.33

8535 Elect.+rel. equip.:insta.+repair.:n.e.c. 52.85

8536 I.t.g.+s.:fabric.,assam.,inst.+rep:el.,

electron.+rel.eg. 42.52

8537 Radio and television repairers 43.76

8538 Labour.+oth.elem.:fab.,ass.,i.,+r.:el.

electron.+rel.eg. 29.59

8539 Fab.,assemb.i.+r.:electric.,electron.+ rel.

equip.:n.e.c. 34.62

8540 Forem./w:fabri.,assam.+rep.:wood prod 39.87

8541 Cabinet and wood furniture makers 32.57

8546 I.t.g.+s.:fab.,ass.+repair.wood prod. 31.98

8548 Labour.+oth.elem.:fab.,assam.,

+repair: wood products 27.61

8549 Fab.,assam.+repair.:wood prod.:n.e.c. 29.04

8550 Forem./w.:fab.,assam.+repair.:textile,

fur+leather prod. 34.53

8551 Patternmaking,marking+cutting:textile

fur+leather prod. 30.32

8553 Tailors and dressmakers 28.52

8555 Furriers 28.91

8557 Milliners,hat and cap makers 22.71

8561 Shoemaking and repairing 25.37

8562 Upholsterers 31.22

8563 Sewing mach. oper.:textile+similar mat. 25.00

8566 Itg.+s.:fabric.,assam.,+repair:textile,

fur+leather 26.78

8568 Labour.+oth.elem.:fab.,assam,+repair:

text.,fur+leather 24.81

8569 Fabric.assam.+repair.:text.,fur+leath. 26.36

8570 Foremen/w:fab.,assam.+repair.:rubber,

plastics+rel. 42.59

8571 Bond.,cement.:rubber,plastics+rel. prod 33.27

8573 Moulding:rubber,plastics+ rel. prod. 30.45

8575 Cut., finish.rubber,plastics+rel. prod. 31.37

8576 I.t.g.+s.:fabric.,assam.+repair.:rubber,

plastics+rel. 36.98

8578 Labour.+oth.elem.:fab.assam.+rep.: rubber,

plastics+rel. 30.37

8579 Fab.,assam.+rep.:rubber,plas.+rel.n.e.c. 31.23

8580 Foremen/w:mechan.+repairers:n.e.c. 48.51

8581 Motor vehicle:mechanics and repairers 39.19

8582 Aircraft:mechanics and repairers 49.42

8583 Rail transport equip.:mechan.+repair. 42.57

8584 Indus.,farm+constr. mach.:mechan.+rep. 46.70

8585 Bus.,comm. mach.:mechan.+rep. 48.13

8586 I.t.g.+s.:equipment repair:n.e.c. 43.87

8587 Watch and clock:repairers 39.87

8588 Precision instrument:make.+repairers 53.83

8589 Other mechanics and repairers:n.e.c. 38.25

8590 Forem./w:oth prod:fab.,ass.+rep.:n.e.c. 42.99

8591 Jewelry,silverw.:fabric.,assam.+repair. 33.35

8592 Marine craft:fabricating,assam.+repair. 37.66

8593 Paper product:fabricating + assembling 32.93

8595 Painting and decorating:n.e.c. 33.30

8596 I.t.g.+s.:other prod. fabric.,assam.

+repair. 33.38

8598 Labour.+oth.elem.:oth. prod.:fab.,

assam.+repair. 30.01

8599 Oth. prod.:fabricat.,assam.+repair.n.e.c. 30.36

8710 Foremen/w:excavat.,grading,paving+rel. 42.54

8711 Excavating,grading and related 35.29

8713 Paving,surfacing and related 30.71

8715 Railway section and track workers 32.64

8718 Lab.+oth.elem.:excav.,grad.,pav.+rel. 28.33

8719 Excav.,grading,paving and rel.:n.e.c. 37.36

8730 Forem./w:el.pow.,light.+wire com. eg. erecting,i.+rep. 57.39

8731 Electrical power line workers and rel. 51.09

8733 Construction electrician and repair. 47.94

8735 Wire comm.+rel. equip.:install.+rep. 50.71

8736 I.t.g.+s.:el.power,light.+wire comm

eg.erecting,i.+rep. 53.53

8738 Labour.+oth.el.:el.power,light.+wire

comm.eg.:er.i.+rep. 36.61

8739 El.power,light.+wire comm.eg.: erecting,

ins.+rep: n.e.c. 47.31

8780 Foremen/women:other constr. trades 44.75

8781 Carpenters and related 34.86

8782 Brick and stone masons+tile setters 36.21

8783 Concrete finishing and related 33.46

8784 Plasterers and related 34.15

8785 Painters,paperhangers and related 31.94

8786 Insulating:construction 34.34

8787 Roofing,waterproofing and related 29.83

8791 Pipefitting,plumbing and related 45.04

8793 Structural metal erectors 40.78

8795 Glaziers 35.07

8796 I.t.g.+s.:other construction trades 48.79

8798 Lab.+oth. elemen.:oth.const. trades 28.13

8799 Other construction trades:n.e.c. 33.43

9110 Foremen/w:air transport operating 58.01

9111 Air pilots,navigat.+flight engineers 64.07

9113 Air transport operating support 53.64

9119 Air transport operating:n.e.c. 45.16

9130 Foremen/w:railway transp. operat. 48.23

9131 Locomotive operating 49.25

9133 Conductors+brake workers:railway 44.28

9135 Railway transp. operating support 42.87

9139 Railway transp. operating:n.e.c. 37.35

9151 Deck officers:ship 56.36

9153 Engineering officers:ship 55.32

9155 Deck crew:ship 36.31

9157 Engine and boiler-room crew:ship 38.48

9159 Water transport operating:n.e.c. 37.15

9170 Foremen/w:motor transport oper. 40.79

9171 Bus drivers 34.93

9173 Taxi drivers and chauffeurs 30.92

9175 Truck drivers 34.45

9179 Motor transport operating:n.e.c. 36.04

9190 Foremen/w:oth. transp. equip. oper. 47.31

9191 Subway+street railway operating 45.62

9193 Rail vehicle oper.,exc. rail transport 40.79

9199 Other transport equip. oper.:n.e.c. 31.93

9310 Foremen/w:mat. handl.+rel.:n.e.c. 42.33

9311 Hoisting:n.e.c. 40.73

9313 Longsho. work.,stevedo.+frei. handl. 32.59

9314 Parcel carriers:n.e.c. 21.86

9315 Material handling equip. oper.:n.e.c. 35.21

9317 Packaging:n.e.c. 25.79

9318 Labour.+oth. elem.:mat. handl.+rel. 28.56

9319 Other material handling+rel.:n.e.c. 31.99

9510 Foremen/women:printing+related 46.36

9511 Typesetting and composing 42.35

9512 Printing press 40.66

9513 Stereotyping and electrotyping 36.43

9514 Print.,engrav.,exc. photo-engraving 48.79

9515 Photo-engraving and related 44.92

9517 Bookbinding and related 30.30

9518 Labouring+other elemental:printing+ rel. 26.37

9519 Printing and related:n.e.c. 31.69

9530 Forem./w.:stat. engine+util.eg.oper.+rel. 56.59

9531 Power station operators 54.46

9539 Station. eng.+util. eg.:oper.+rel.:n.e.c. 47.63

9550 Forem/w.:elect.+rel.com.eg.:oper.:n.e.c. 57.85

9551 Radio+tel. broadcas. equip. operators 50.27

9553 Telegraph operators 44.38

9555 Sound+video record.+repr. equip. oper 49.49

9557 Motion picture projectionists 43.65

9559 Oth. elec.+rel. com.equip.:oper.:n.e.c. 45.78

9590 Foremen/w:oth. crafts+equ.:oper.:n.e.c. 50.82

9591 Photographic processing 37.19

9599 Other crafts and equipment:oper.:n.e.c. 44.12

9910 Supervisors and foremen/women:n.e.c. 48.27

9916 Inspect.,test.,grading+sampling:n.e.c. 42.68

9919 Oth. occup.:not elsewhere classified 34.90

9921 Labour.+other elemental:manu. 28.97

9922 Labour.+oth. elemental:trans.+comm. 31.28

9923 Labour.+other elemental:trade 23.41

9924 Labour.+other elemental:service 21.26

9925 Labour.+oth. elem.:public admin.+def. 26.16

9926 Labour.+other elemental:other indus. 24.11

4.13 Map of Variables

Campaign-Period: Front End Variables

Name Label

IDNUM Respondent Identification

Number

PROVINCE Province Of Interview

AREACODE Telephone Area Code

CPSSAMP Day Of Sample Release

CPSREP Sample Subsets <Replicate>

CPSTIME Length Of Interview <Minutes>

CPSRES Outcome Of Interview

NADULTS Number Of Adults <Cdn

Citizens> In Hhld

CPSATTEM Total Number Of Call Attempts

CPSCONT Total Times Respondent Contacted

CPSANS Number Of Times Telephone

Answered

CPSREFUS Number Of Refusals Before

Completion

CPSDATE Date Of Interview <MDDYY>

CPSINUM Interviewer's Number (# Of

Interviews)

CPSIGEN Interviewer's Gender (# Of

Interviews)

CPSLANG Language Of Interview

CPSRN1 Question Order/Random Delivery

<A1A-A2H>

CPSRN2 Question Order/Random Delivery

<C3-C5>

CPSRN3 Question Order/Random

Delivery<F10A-F10G>

CPSRN4 Route Country/Riding Split

<I1A--I2D>

CPSRN5 Randomize Party Leaders

<D1A-D1C>

CPSRN6 Randomize Federal Parties

<D1G-D1I>

CPSRN7 Randomize Party Leaders

<H1A--H4E>

CPSRN8 Randomize Parties In Riding

<I1A-I1C>

CPSRN9 Randomize Parties In Country

<I2A-I2D>

CPSRN10 Question Order/Random Delivery

<J1A-J1G>

CPSRGEN Respondent's Gender

CAMPAIGN PERIOD

SECTION A: VOTE INTENTIONS

Name Label

CPSJF1 Timer - Response Time Item

<CPSA4>

CPSA1 Most Important Issue To You Personally

CPSA2A Personally> Preserving National Unity

CPSA2B Personally> Reducing The Deficit

CPSA2C Personally> Creating Jobs

CPSA2D Personally> Cutting Taxes

CPSA2E Personally> Keeping Election Promises

CPSA2F Personally> Protecting Social Pgms

CPSA2G Personally> Fighting Crime

CPSA2H Personally>Defending Interests Of

Quebec

CPSA3 On Election Day Certain To Vote

CPSA4 Party Think Will Vote For <CPSJF1>

CPSA5 Don't Know,NoParty> Party Leaning

Toward

CPSA7 Which Party Would Be Your Second

Choice

CAMPAIGN PERIOD

SECTION B: INTEREST AND CAMPAIGN ACTIVITIES

Name Label

CPSB1 Attention Paid>News About Election On

CPSB2 Attention Paid>ElectionNews In

Newspaper

CPSB3 Attention Paid>ElectionNews On The

Radio

CPSB4 How Rate Interest In This Election

CPSB5 How Rate Interest In Politics Generally

CPSB6 See TV Commercials For A Political

Party

CPSB7 Hear RadioCommercials For

PoliticalParty

CPSB8 Talked About Election With

Friends/Rels

CPSB8A Talked About Election With Other

People

CPSB8B Disagreed With People You Talked

With

CPSB9 Satisfaction>Way Democracy

WorksInCanada

CPSB10A Elected To Parliament Lose Touch

People

CPSB10B People Like Me NotHave Say What

Gov Does

CPSB10C Politics&Government Seem So

Complicated

CPSB10D Not Think Gov't Cares What People

Think

CPSB10E Politicans Ready To Lie To Get

Elected

CAMPAIGN PERIOD

SECTION C: PERSONAL FINANCES AND THE ECONOMY

Name Label

CPSC1 Better/Worse Off Financially Than Yr

Ago

CPSC1A Much/Somewhat Better Off Than Year

Ago

CPSC1B Much/Somewhat Worse Off Than Year

Ago

CPSC2 Better/Worse Off Financially Yr From

Now

CPSC2A Much/Somewhat Better Off Year From

Now

CPSC2B Much/Somewhat Worse Off Year From

Now

CPSC3 Policies Of Federal Government Made

You:

CPSC4 Policies Provincial Government Made

You:

CPSC5 Unemployment Since Liberals

CameToPower

CPSC6 Next Few Years Unemployment Will

Go Up

CAMPAIGN PERIOD

SECTION D: LEADER AND PARTY EVALUATION

Name Label

CPSDR1 How Much Know About> Jean Chrétien

CPSDR2 How Much Know About> Jean Charest

CPSDR3 How Much Know About> Alexa

McDonough

CPSDR4 How Much Know About> Preston Manning

CPSDR5 How Much Know About> Gilles Duceppe

CPSD1A Rating> Jean Charest

CPSD1B Rating> Jean Chrétien

CPSD1C Rating> Alexa McDonough

CPSD1D Rating> Preston Manning

CPSD1E Rating> Gilles Duceppe <Que Only>

CPSD1G Rating> Federal Conservative Party

CPSD1H Rating> Federal Liberal Party

CPSD1I Rating> Federal New Democratic Party

CPSD1J Rating> Reform Party

CPSD1K Rating> Bloc Quebecois

CPSD1L Rating> Politicians In General

CPSD1F Rating> Brian Mulroney

CPSD1N Rating> Lucien Bouchard

CAMPAIGN PERIOD

SECTION E: PLACEMENTS

Name Label

CPSE1A Cut Taxes=CutSocialPgms/

Increase=Improve

CPSE1B Cut Taxes> By How Much

CPSE1C Increase Taxes> By How Much

CPSE1D Liberal Party Wants To Cut Taxes

CPSE1E Liberals Cut Taxes> By How Much

CPSE1F Liberals Increase Taxes> By How Much

CPSE1G Conservatives Want To Cut Taxes

CPSE1H Conservatives Cut Taxes> By How Much

CPSE1I Conservatives Increase Taxes>By How

Much

CPSE1J New Democratic Party Wants To Cut Taxes

CPSE1K NDP Cut Taxes> By How Much

CPSE1L NDP Increase Taxes> By How Much

CPSE1M Reform Party Wants To Cut Taxes

CPSE1N Reform Party Cut Taxes> By How Much

CPSE1O Reform Party Increase Taxes> By How

Much

CPSE1P Bloc Quebecois Wants To Cut Taxes

CPSE1Q Bloc Quebecois Cut Taxes> By How

Much

CPSE1R BlocQuebecois Increase Taxes>By

How Much

CPSE3A How Much Should Be Done For

Quebec

CPSE3B More Done For Quebec> How Much

CPSE3C More Done For Quebec> How Much

Less

CPSE3D How Much Liberals Want To Do For

Quebec

CPSE3E Liberals Do For Quebec> How Much

CPSE3F Liberals Do For Quebec> How Much

Less

CPSE3G How Much Conservatives WantDo For

Quebec

CPSE3H Conservatives DoForQuebec> How

Much More

CPSE3I Conservatives DoForQuebec> How

Much Less

CPSE3J How Much NDP Want To Do For

Quebec

CPSE3K NDP Do For Quebec> How Much

CPSE3L NDP Do For Quebec> How Much Less

CPSE3M How Much Reform Party WantsDo

For Quebec

CPSE3N Reform Party Do ForQuebec> How

Much More

CPSE3O Reform Party Do ForQuebec> How

Much Less

CPSE3P How Much BlocQuebecois WantDo

For Quebec

CPSE3Q BlocQuebecois DoForQuebec> How

Much More

CPSE3R BlocQuebecois DoForQuebec> How

Much Less

CAMPAIGN PERIOD

SECTION F: POLICY I

Name Label

CPSF1 How Much ShldBeDoneFor Racial

Minorities

CPSF2 Only Married People Be Having Children

CPSF3 Better Off Women StayedHome

WithChildren

CPSF4 Not Much Any Gov DoTo Solve

Unemployment

CPSF5 Maintain Social Pgms=Eliminate Deficit

CPSF6 Gov Leave To Private Sector Create Jobs

CPSF7 Liberals Elected To Do In 1993

CPSF8 Best Way To Fight Unemployment

CPSF10A Liberals> Preserving National Unity

CPSF10B Liberals> Reducing The Deficit

CPSF10C Liberals> Creating Jobs

CPSF10D Liberals> Keeping Election Promises

CPSF10E Liberals> Defending Interests Quebec

CPSF10F Liberals> Fighting Crime

CPSF10G Liberals> Protecting Social Programs

CPSF11 Job By Reform Party In Parliament

CPSF12 Job By Bloc Quebecois In Parliament

CPSF13 Party PromisingTo Lower Income Taxes

10%

CPSF14 Promising Cut UnemploymentInHalf By

2001

CPSF15 Party Against Quebec As Distinct Society

CAMPAIGN PERIOD

SECTION G: NATIONAL ECONOMIC CONDITIONS

Name Label

CPSG1 Over The Past Year Canada's Economy

CPSG1A Federal Policies=Canada's Economy Better

CPSG1B Federal Policies=Canada's Economy Worse

CPSG2 Over The Past Year <Prov>'s Economy

CPSG2A Federal Policies <Prov>'s Economy Better

CPSG2B Federal Policies <Prov>'s Economy Worse

CPSG3A Next 12 Months Canada's Economy

CPSG3B Next 12 Months <Prov>'s Economy

CAMPAIGN PERIOD

SECTION H: LEADER TRAITS - RANDOMIZE ORDER OF LEADERS

Name Label

CPSH1A Describe> Charest> Strong Leader

CPSH1B Describe> Charest> Trustworthy

CPSH1C Describe> Charest> Arrogant

CPSH1D Describe> Charest> Compassionate

CPSH1E Describe> Charest> In Touch With

Times

CPSH2A Describe> Chrétien> Strong Leader

CPSH2B Describe> Chrétien> Trustworthy

CPSH2C Describe> Chrétien> Arrogant

CPSH2D Describe> Chrétien> Compassionate

CPSH2E Describe> Chrétien> In Touch With

Times

CPSH3A Describe> McDonough> Strong Leader

CPSH3B Describe> McDonough> Trustworthy

CPSH3C Describe> McDonough> Arrogant

CPSH3D Describe> McDonough>

Compassionate

CPSH3E Describe> McDonough> In Touch

With Times

CPSH4A Describe> Manning> Strong Leader

CPSH4B Describe> Manning> Trustworthy

CPSH4C Describe> Manning> Arrogant

CPSH4D Describe> Manning> Compassionate

CPSH4E Describe> Manning> In Touch With Times

CPSH5A Describe> Duceppe> Strong Leader

CPSH5B Describe> Duceppe> Trustworthy

CPSH5C Describe> Duceppe> Arrogant

CPSH5D Describe> Duceppe> Compassionate

CPSH5E Describe> Duceppe> In Touch With

Times

CAMPAIGN PERIOD

SECTION I: PARTY CHANCES - RANDOMIZE ORDER OF PARTIES

Name Label

CPSI1A PC Chances> Winning In Your Riding

CPSI1B Lib Chances> Winning In Your Riding

CPSI1C NDP Chances> Winning In Your

Riding

CPSI1D Reform Chances> Winning In Your

Riding

CPSI1E Bloc Chances> Winning In Your

Riding

CPSI2A PC Chances> Winning In Whole

Country

CPSI2B Lib Chances> Winning In Whole

Country

CPSI2C NDP Chances> Winning In Whole

Country

CPSI2D Reform Chances> Winning In Whole

Country

CPSI2E Bloc Chances>Majority Of Seats In Quebec

CPSI3A Lib Chances> Forming Offical Opposition

CPSI3B PC Chances> Forming Official Opposition

CPSI3C NDP Chances> Forming Official

Opposition

CPSI3D Reform Chances> Form Official

Opposition

CPSI3E Bloc Chances>Forming Official Opposition

CAMPAIGN PERIOD

SECTION J: POLICY II

Name Label

CPSJ1A Party Best At> Preserving National Unity

CPSJ1B Party Best At> Creating Jobs

CPSJ1C Party Best At> Cutting Taxes

CPSJ1D Party Best At> Keeping Promises

CPSJ1E Party Best At>Defending Interests Quebec

CPSJ1F Party Best At> Protecting Social Pgms

CPSJ1G Party Best At> Fighting Crime

CPSJ2 1993 Campaign> Liberals Promise No GST

CPSJ2B Liberals Really Try Keep GST Promise

CPSJ2C Liberals Not Try Keep Promise, How Feel

CPSJ3 Quebec Be Recognized As Distinct Society

CPSJ3C Change Mind If Distinct Keeps Quebec Can

CPSJ3A Favourable To Quebec Sovereignty <Que>

CPSJ4 How Likely That Quebec Will Separate

CPSJ4A Possibility Of Separation Worry You

CPSJ5 Gap Between Rich And Poor In Canada

CPSJ5A Gap Between Rich And Poor Increased

CPSJ6 Federal Spending Cuts Been Fair/Unfair

CPSJ7 Who Has Been Hardest Hit By

SpendingCuts

CPSJ9 Aboriginal Peoples Compared Other Cdns

CPSJ10 Federal Spending For Aboriginal Peoples

CPSJ18 Canada Should Admit More Immigrants

CPSJ19 Think Pollution In Canada Has Got Worse

CPSJ20 Think That Crime In Canada Has Gone Up

CPSJ21 Deal With Young Offenders=Violent Crime

CPSJ12 Federal Gov Treat <Prov> As Other Parts

CPSJ13 Political Parties Keep Election Promises

CPSJ14 Have You Heard About The Somalia Affair

CPSJ14A How Federal Gov't Handled Somalia Affair

CPSJ15 Gov's Decision Hold Election AtThis Time

CAMPAIGN PERIOD

SECTION K: PARTY IDENTIFICATION AND VOTE HISTORY

Name Label

CPSK1 Federal Politics Think Of Self

As<Party>

CPSK2 How Strongly <Federal Party> Do You

Feel

CPSK3 A Little Closer To One Federal Party

CPSK4 Which Federal Party Closer To

CPSK5 Vote In Last Federal Election - 1993

CPSK6 Party Voted For> Last Federal Election

CPSK13 Provincial Election Held Today,Vote

For:

CPSK14 Provincial Party Leaning Toward

CPSK15 Vote In Quebec Referendum On

Sovereignty

CPSK16 Vote Yes Or No In 1995 Quebec

Referendum

CAMPAIGN PERIOD

SECTION L: DEBATE

Name Label

CPSL1 See English TV Debate Among

PartyLeaders

CPSL1A See All Of The English TV Debate

CPSL1B Which Party Leader Did Best In TV

Debate

CPSL1C Which PartyLeader Did Worst In TV

Debate

CPSL2 See French TV Debate Among Party

Leaders

CPSL2A See All Of The French TV Debate

CPSL2B Which Party Leader Did Best In TV

Debate

CPSL2C Which PartyLeader Did Worst In TV

Debate

CPSL3 Past Week> See/Hear Polls How Well

Doing

CPSL4 Where Get Most Information Re

Election

CPSL5 Most Important Source Of Election

Info

CPSL6 Recall The Name Of The President Of

USA

CPSL11 Recall Name> Federal Minister Finance

CPSL12 Recall Name> Premier Of <Province/Terr>

CPSL13 Recall Name> First Woman PM Of Canada

CAMPAIGN PERIOD

SECTION M: BACKGROUND

Name Label

CPSAGE Respondent's Year Of Birth

CPSM2 Respondent's Marital Status

CPSM3 Highest Level Of Education Completed

CPSM4 Present Employment Status

CPSM4A <If CPSM4=4,5,6,7> Main Income

Earner

CPSM6 Occupation <Stats Canada 1980 CCDO

Code>

CPSM7 Work For Private Firm/Public/

Government

CPSM7A Work For Federal/Provincial Government

CPSM8 Out Of Work/Laid Off During Last Year

CPSM8A Worried About Job In The Near Future

CPSM9 Do You/Hhld Member Belong To A

Union

CPSM10 Religious Affiliation

CPSM10A Church Or Denomination

CPSM11 Respondent's Country Of Birth

CPSM12 Year Come To Live In Canada

CPSM13 To What Ethnic Or Cultural Group

Belong

CPSM14 Language Usually Speak At Home

CPSM15 Language First Learned&Still Understand

CPSM16 Total Household Income <Thousands>

CPSM16A Total Household Income <Category>

CPSM17 # Of Children Under 18 Live In Home

CPSM18 R Have Long-Term Disability/Handicap

CPSM19 Long-Term Disability/Handicap Affect:

CPSKNOW R's General Level Of Knowledge&Info

POSTCODE Postal Code <Forward Sortation Area>

BLISH81R Occupation:Respondent> Blishen 1981

SES

PINPORR Respondent> Pineo-Porter 1981 Category

WEIGHT VARIABLES

Name Label

CPSHHWGT Household Weight - CES Campaign

CPSPWGT1 Provincial Weight <All> - CES

Campaign

CPSPWGT2 Provincial Weight <No PQ> - CES

Campaign

CPSNWGT1 National Weight <All> - CES

Campaign

CPSNWGT2 National Weight <No PQ> - CES

Campaign

POST ELECTION: FRONT END VARIABLES

Name Label

PESSAMP Day Of Sample Release

PESREP Sample Subsets <Replicate>

PESTIME Length Of Interview <Minutes>

PESRES Outcome Of Interview

PESATTEM Total Number Of Call Attempts

PESCONT Total Times Respondent Contacted

PESANS Number Of Times Telephone

Answered

PESREFUS Number Of Refusals Before

Completion

PESDATE Date Of Interview <MMDDYY>

PESINUM Interviewer's Number (# Of

Interviews)

PESIGEN Interviewer's Gender (# Of

Interviews)

PESLANG Language Of Interview

PESRN12 Question Order/Random Delivery

<A2A,A2B>

PESRN2 Question Order/Random Delivery

<A4C,A4D>

PESRN10 Question Order/Random

Delivery<E5A--E5C>

PESRN11 Question Order/Random

Delivery<E7A,E7B>

PESRN4 Question Order/RandomDelivery

<F14A,F14B>

PESRN16 QuestionOrder/RandomDelivery <H1-5H10-14>

PESRN14 Question Order/Random

Delivery<I5A--I5F>

PESRN5 Randomize Party Leaders

<C1A--C1E>

PESRN6 Randomize Federal Parties

<C2A--C2E>

PESRN8 Randomize Parties In Riding

<C3A--C3E>

PESRN9 Question Order/Random

Delivery<E6A--E6G>

PESRN3 Question Wording <E11B>

PESRGEN Respondent's Gender

POST ELECTION

SECTION A: THE VOTE

Name Label

PESA1 Most Important Issue To You Personally

PESA2A Did You Vote In The Election

PESA2B <Democracy> Did You Vote In Election

PESA3 Particular Reason Why You Did Not Vote

PESA4 Which Party Did You Vote For

PESA4A Your Preference For The <Voted> Party

PESA4B Which Party Was Your Second Choice

PESA4C When Decide You Were GoingTo

Vote<Party>

PESA4D When Decide You Were GoingTo

Vote<Party>

PESA5B Satisfaction> Way Democracy Works

Canada

PESA6 Having Reform As The Official

Opposition

PESA7A Conservatives And Reform Joined

Together

PESA8 What Have The Liberals Been Elected

ToDo

POST ELECTION

SECTION B: INTEREST AND MEDIA

Name Label

PESB1 Attention Paid>News About Election On

PESB2 Attention Paid>ElectionNews In

Newspaper

PESB3 Attention Paid>ElectionNews On The

Radio

PESB4 How Rate Interest In Election Campaign

PESB5 Talked About Election With Friends/Rels

PESB6 Talked About Election With Other People

POST ELECTION

SECTION C: LEADER, PARTY, CANDIDATE EVALUATION

Name Label

PESDR1 How Much Know About> Jean

Chrétien

PESDR2 How Much Know About> Jean

Charest

PESDR3 How Much Know About> Alexa

McDonough

PESDR4 How Much Know About> Preston

Manning

PESDR5 How Much Know About> Gilles

Duceppe

PESFLAG1 Party Leader Randomization

PESC1A Rating> Jean Charest

PESC1B Rating> Jean Chrétien

PESC1C Rating> Alexa McDonough

PESC1D Rating> Preston Manning

PESC1E Rating> Gilles Duceppe <Que Only>

PESFLAG2 Political Party Randomization

PESC2A Rating> Federal Conservative Party

PESC2B Rating> Federal Liberal Party

PESC2C Rating> Federal New Democratic

Party

PESC2D Rating> Reform Party

PESC2E Rating> Bloc Quebecois

PESC2F Rating> Politicians In General

PESC3A Rating> Conservative Candidate

PESC3B Rating> Liberal Candidate

PESC3C Rating> NDP Candidate

PESC3D Rating> Reform Candidate <ROC

Only>

PESC3E Rating> Bloc Candidate <Que Only>

PESC4 Rating> Paul Martin

PESC5 Rating> Provincial Premier

PESC6 Rating> Pierre Elliott Trudeau

POST ELECTION

SECTION E: POLICY

Name Label

PESE6A Cut Spending> Defence

PESE6B Cut Spending> Welfare

PESE6C Cut Spending> Pensions/Old Age

Security

PESE6D Cut Spending> Health Care

PESE6E Cut Spending> Unemployment Insurance

PESE6F Cut Spending> Education

PESE6G Cut Spending> Aid Developing Countries

PESE3 How Much Power Think Unions Should

Have

PESE2 How Much Should Be Done For Business

PESE1 How Much Should Be Done For Women

PESE4 Canada's Ties With The United States

PESE5A Opinion> 3 Positions: Abortion

PESE5B Opinion> 3 Positions: Abortion

PESE5C Opinion> 3 Positions: Abortion

PESE7A Opinion> Government Services

PESE7B Opinion> Government Services

PESE9A Last Five Years, Quality Of Education

PESE9B Quality Of Education How Much Worse

PESE9C Last Five Years, Quality Of Health Care

PESE9D Quality Of Health Care How Much Worse

PESE9E Most Responsible For Cuts To Health

Care

PESE10 Favourable To Quebec Sovereignty

<Que>

PESE11B Likely Canada Become Part Of U.S.

<ROC>

PESE10A Standard Of Living If Quebec Separates

PESE10B Standard Of Living How Much Better

PESE10C Standard Of Living How Much Worse

PESE10D French Language Threatened In Quebec

PESE10E Que Separates, French Language In

Quebec

PESE10F Language Situation How Much Better

PESE10G Language Situation How Much Worse

PESE11D Que Separates,Close EconomicUnion

Canada

PESE12 Only Police & Military Allowed Have

Guns

PESE13 Capital Punishment Is Never Justified

PESE15 Politicians Ready To Lie To Get Elected

PESE16 Gov Accepts High Level

Unemploy=Defeated

PESE18 Everyone Should Be Forced Retire At 65

PESE28 Issues That Matter To Women

NotDiscussed

PESE25 Good Thing Canada&USA Become

One Country

PESE19 Not Get Ahead, Blame Self, Not System

PESE20 Businesses Make Money, Everyone

Benefits

PESE21 Trust Down-To-Earth People Than

Experts

PESE22 Party Promised Lower Personal Tax 10

PESE23 Party Promised Cut Unemployment In

Half

PESE29 Party Said All Provinces Treated

Equally

PESE30 What Meant By Provinces Treated

Equally

PESE24 Party Against Quebec As Distinct

Society

POST ELECTION

SECTION F: COLLECTIVITIES

Name Label

PESF1 Rating> How Feel About Big Business

PESF2 Rating> How Do You Feel About

Unions

PESF3 Rating> How Do You Feel About

Feminists

PESF5 Rating> How Feel About People On

Welfare

PESF6 Rating>How Feel About Aboriginal

Peoples

PESF7 Rating> How Do You Feel About The

Police

PESF8 Rating> How Feel About Racial

Minorities

PESF9 Rating>How Do You Feel About

Babyboomers

PESF10 Rating> How Feel About Gays &

Lesbians

PESF11A Rating> How Do You Feel About

Canada

PESF11B Rating> How Do You Feel About

Province

PESF12 Rating> How Do You Feel About

Quebec

PESF13 Rating> How Feel About The United

States

PESF14A Government Looks After People

PESF14B Government Looks After People

POST ELECTION

SECTION H: PARTY IDENTIFICATION AND VOTE HISTORY

Name Label

PESH1 Federal Politics Think Of Self As<Party>

PESH2 How Strongly <Federal Party> Do You Feel

PESH3 A Little Closer To One Federal Party

PESH4 Which Federal Party Closer To

PESH5 Federal Politics Close To Any Party

PESH6 Which Federal Party Close To

PESH7 Do You Feel How Close To <Federal Party>

PESH8 Feel A Little Closer To One Party

PESH9 Which Party Feel A Little Closer To

PESH21 Federal Political Party Too Extreme

PESH10 Prov Politics Think Of Self As <Party>

PESH11 How Strongly <Prov Party> Do You Feel

PESH12 A Little Closer To One Provincial Party

PESH13 Which Provincial Party Closer To

PESH14 Provincial Politics Close To Any Party

PESH15 Which Provincial Party Close To

PESH16 Do You Feel How Close To <Prov Party>

PESH17 Feel A Little Closer To One Party

PESH18 Which Party Feel A Little Closer To

PESH19 Vote In The Last Provincial Election

PESH20 Party Voted For>Last Provincial Election

PESA5 Vote Federal Elections If Unable On Day

PESA5A How Vote Fed Election If Unable On Day-1

PESA5A2 How Vote Fed Election If Unable On Day-2

PESA5A3 How Vote Fed Election If Unable On Day-3

POST ELECTION

SECTION I: MEDIA HABITS

Name Label

PESI1 # Hours A Day Usually Watch TV

PESI2 Do You Watch Canadian Stations

PESI3 # Hours A Day Listen To The Radio

PESI4 # Days A Week Read A Newspaper

PESB7 See English TV Debate Among PartyLeaders

PESB7A See All Of The English TV Debate

PESB7B Which Leader Did Best In English Debate

PESB7C Which Leader Did Worst In English

Debate

PESB8 See First French TV Debate (May 13)

PESB8A See All Of First French TV Debate

PESB8B Who Did The Best In First French Debate

PESB8C Who Did The Worst In First French

Debate

PESB9 See Second French TV Debate (May

19)

PESB9B Who Did The Best In Second French

Debate

PESB9C Who Did The Worst In Second

FrenchDebate

PESE21A Chrétien Betrayed Quebec

Constitutional

PESE21C Reform Party Only Speaks For The

West

PESE21D Charest Has Style, But Not Much To

Say

PESE21E Manning Is Threat To Canadian Unity

PESE21F NDP Is Out Of Touch With The Times

PESE21G No Reason Sovereignist Party In

Ottawa

PESE21I Best Way Defend West=Elect Reform

Mps

PESE21J Jean Charest Was A One Man Show

PESE21K Jean Charest Too Close To Brian

Mulroney

PESE27 QueSeparates,Canada Close Economic

Union

PESI5A Four Goals> Most Important

PESI5B Four Goals> Next Most Important To

You

PESI5C Four Goals> Least Most Important To

You

PESI5D Four Goals> Most Important

PESI5E Four Goals> Next Most Important To

You

PESI5F Four Goals> Least Most Important To

You

POST ELECTION

SECTION M: BACKGROUND

Name Label

PESAGE Respondent's Year Of Birth

PESM10B In Your Life, Importance Of Religion

PESM12 Respondent's Province/Territory Of Birth

PESM20 # Of Years Lived In

MAILBACK QUESTIONNAIRE

SECTION A

Name Label

MBSQLANG Language Of Questionnaire

MBSA1 Gone Too Far In Pushing Equal Rights

MBSA2 Be More Tolerant People Choose

Standards

MBSA3 Lay Off Women Whose Husbands Have

Jobs

MBSA4 Gov Do More Reduce Income Gap

Rich&Poor

MBSA5 Difficult Women Get Jobs Equal

Abilities

MBSA6 Protect Env. More Imp Than Creating

Jobs

MBSA7 NewerLifestyles Contrib

BreakdownSociety

MBSA8 Change=Adapt Our View Of Moral

Behaviour

MBSA9 Fewer Problems=Traditional Family

Values

MBSA10 Not Big Problem Some Have More

Chance

MBSA11 Look After Cdns BornHere

First,Others2nd

MBSA12 People Really Want Work, Can Find A

Job

MBSA13 Tough For Young Because Of

Babyboomers

MBSA14 Minority Groups Need Special Rights

MBSA15 DoMore Protect Cdn Business From

Foreign

MAILBACK QUESTIONNAIRE

SECTION B

Name Label

MBSB1 Government Should <Standard Of

Living>:

MBSB2 Government Should

<Environment>:

MBSB3 Workers And Management

<Conflict>:

MBSB4 When It Comes To Job Hiring

<Quotas>:

MBSB5 Closer To Your View <People On

Welfare>:

MBSB6 People In Government <Waste Tax

Money>:

MAILBACK QUESTIONNAIRE

SECTION C

Name Label

MBSC1A Influence Has> Unions

MBSC2A Influence Has> Farmers

MBSC3A Influence Has> Big Business

MBSC4A Influence Has> Media

MBSC5A Influence Has> Public Sector

Workers

MBSC6A Influence Has> Banks

MBSC7A Influence Has> Consumers

MBSC8A Influence Has> Feminists

MBSC9A Influence Has> Aboriginal Peoples

MBSC10A Influence Has> Racial Minorities

MBSC11A Influence Has> People On Welfare

MBSC12A Influence Has> Small Business

MBSC13A Influence Has> Senior Citizens

MBSC14A Influence Has> Gays And Lesbians

MBSC15A Influence Has> Babyboomers

MBSC16A Influence Has> Environmentalists

MBSC1B Influence ShldHave> Unions

MBSC2B Influence ShldHave> Farmers

MBSC3B Influence ShldHave> Big Business

MBSC4B Influence ShldHave> Media

MBSC5B Influence ShldHave>Public Sector

Workers

MBSC6B Influence ShldHave> Banks

MBSC7B Influence ShldHave> Consumers

MBSC8B Influence ShldHave> Feminists

MBSC9B Influence ShldHave> Aboriginal Peoples

MBSC10B Influence ShldHave> Racial Minorities

MBSC11B Influence ShldHave> People On Welfare

MBSC12B Influence ShldHave> Small Business

MBSC13B Influence ShldHave> Senior Citizens

MBSC14B Influence ShldHave> Gays And Lesbians

MBSC15B Influence ShldHave> Babyboomers

MBSC16B Influence ShldHave> Environmentalists

MAILBACK QUESTIONNAIRE

SECTION D

Name Label

MBSD1 Most People Not Know What Best For

Them

MBSD2 People Have Sense Tell Gov't Do Good

Job

MBSD3 Solve National Prob=GrassRoots

Decisions

MBSD4 Gov ShldPay Most Attention

Well-Informed

MBSD5 All Federal Parties Basically The Same

MBSD6 Parties Spend TooMuch Time Re

Minorities

MBSD7 Gone Too Far Pushing Bilingualism

MBSD8 Protect Women's Interests=More In Parlia

MBSD9 Profits Cdn Banks Making Are A Scandal

MBSD10 Unemployed Move To Regions Where

Jobs

MBSD11 International Trade Creates More Jobs

MBSD12 Immigrants Make Important Contribution

MAILBACK QUESTIONNAIRE

SECTION E

Name Label

MBSE1 Your Opinion> Treatment Of People:

MBSE2 Your Opinion> The Feminist Movement:

MBSE3 More Important In Democratic Society:

MBSE4 Your View> Equality Of Men & Women:

MBSE5 Law Conflicts Charter, Have Final Say:

MBSE6 Your View> Marital Violence:

MBSE7 Feminist Movement Encourages Women:

MBSE8 Your View> Aboriginal Peoples:

MBSE9 Think That> People Running Gov Crooked

MAILBACK QUESTIONNAIRE

SECTION F

Name Label

MBSF1 Confidence> Organised Religion

MBSF2 Confidence> The Armed Forces

MBSF3 Confidence> Public Schools

MBSF4 Confidence> The Courts

MBSF5 Confidence> The Civil Service

MBSF6 Confidence> Unions

MBSF7 Confidence> The Police

MBSF8 Confidence> The Federal Government

MBSF9 Confidence> Provincial/Terr

Government

MBSF10 Confidence> Big Business

MBSF11 Confidence> The Media

MAILBACK QUESTIONNAIRE

SECTION G

Name Label

MBSG1 Participate Peacekeeping Even If Risk

MBSG2 Respect For Authority Children

ShldLearn

MBSG3 Homosexual Couples Allowed Legally

Marry

MBSG4 Too Many Recent Immigrants Not

Want Fit

MBSG5 Caring For Children, Men Less Patient

MBSG6 Anglos In Que Better Treated Fr In

ROC

MBSG7 Quebec Has Right To Separate

MBSG8 Quebec Separates,Aboriginals Remain

Part

MBSG9 Have Right To Work In Region

Where Born

MBSG10 Free Trade With U.S. Has Been Good

MAILBACK QUESTIONNAIRE

SECTION H

Name Label

MBSH1 Political Parties/Cands Spend

MuchAsWant

MBSH2 YourView>Allowed

AdvertiseDuringCampaign

MBSH3 Limit Individuals/Groups Spending On

Ads

MBSH4 Individuals/Groups Can Advertise, Spend:

MBSH5 Referendums On Important Questions

Held:

MBSH6 Have Referendums Same Time As

Elections

MBSH7 Win Majority Seats W/O Majority Of

Votes

MAILBACK QUESTIONNAIRE

SECTION I

Name Label

MBSI1 Satisfaction Way Democracy Works

Canada

MBSI2 Last Election In Canada Conducted Fairly

MBSI3 Think Of Self As Close To PoliticalParty

MBSI3A Which Political Party Closer To

MBSI3B Feel Very Close To This Political Party

MBSI4 Parties In Canada Care What People Think

MBSI5 Parties Necessary To Make System Work

MBSI6 Name Candidate Ran In Your Riding

MBSI7A Rate> Liberal Party

MBSI7B Rate> Progressive Conservative Party

MBSI7C Rate> New Democratic Party

MBSI7D Rate> Reform Party

MBSI7E Rate> Bloc Quebecois

MBSI8A Rate> Jean Chrétien

MBSI8B Rate> Jean Charest

MBSI8C Rate> Alexa McDonough

MBSI8D Rate> Preston Manning

MBSI8E Rate> Gilles Duceppe

MBSI9 State Of Economy These Days In Canada

MBSI10 Past 12 Mos> State Of Economy In Canada

MBSI11 Members Parliament Know Ordinary

Think

MBSI12 Past 12 Mos> Contact Member Parliament

MBSI13 Makes A Difference Who Is In Power

MBSI14 Who People Vote For Can Make

Difference

MBSI15 People In Canada Say What Think:Politics

MBSI16A Scale> Where Place Self In Politics

MBSI16B Scale> Where Place Liberal Party

MBSI16C Scale> Where Place Conservative Party

MBSI16D Scale> Where Place NDP

MBSI16E Scale> Where Place Reform Party

MBSI16F Scale> Where Place Bloc Quebecois

MAILBACK QUESTIONNAIRE

SECTION J

Name Label

MBSJ1 Respondent's Year Of Birth

MBSJ2 Respondent's Gender

MAILBACK QUESTIONNAIRE

SECTION K

Name Label

MBSK1A Voting From Home By Mail

MBSK1B Voting From Home By Telephone

MBSK1C Voting From Home By Computer

MBSK1D Voting At Station: Touch Computer

Screen

MBSK1E Voting At Station:Ballots

MachineCounted

MBSK2A1 Most Likely> Home By Mail

MBSK2A2 Most Likely> Home By Telephone

MBSK2A3 Most Likely> Home By Computer

MBSK2A4 Most Likely> Touch Computer

Screen

MBSK2A5 Most Likely> Ballots Machine

Counted

MBSK2B1 Least Likely> Home By Mail

MBSK2B2 Least Likely> Home By Telephone

MBSK2B3 Least Likely> Home By Computer

MBSK2B4 Least Likely> Touch Computer

Screen

MBSK2B5 Least Likely> Ballots Machine

Counted

MBSK3A Reason For Your Least Likely

Choice-1st

MBSK3B Reason For Your Least Likely

Choice-2nd

COMMENTS Any Further Comments

SENTQ Respondents To Whom Questionnaire

Mailed

ANALYSIS ASSISTANCE VARIABLES

Name Label

RTYPE1 Respondent <Campaign>

RTYPE2 Respondent <Post Election>

RTYPE3 Respondent <Mailback>

RLINK Linking CPS/PES/MBS Respondents

WAVE Responded To What Wave Of The Survey

References

Babbie, Earl. 1992. The Practice of Survey Research (6th Edition). Belmont, California:

Wadsworth.

Bartels, Larry M. 1988. Presidential Primaries and the Dynamics of Public Choice.

Princeton: Princeton University Press.

Bassili, John N. 1993. "Response Latency versus Certainty." Public Opinion Quarterly, 57 (1) 54-61.

Bassili, John N. 1996. "The How and Why of Response Latency Measurement in

Surveys." In: Answering Questions, eds., Norbert Schwarz and Seymour Sudman. San Francisco: Jossey-Bass.

Bassili, John N. and Joseph F. Fletcher. 1991. "Response-Time Measurement in

Surveys." Public Opinion Quarterly, 55 (3) 331-346.

Blais, André and Martin Boyer. 1996. "Assessing the Impact of Televised Debates: The Case of the 1988 Canadian Election." British Journal of Political Science, 26: 143-164.

Blishen, Bernard, Bill Carroll and Catherine. Moore. 1987. "The 1981 Socioeconomic

Index for Occupations in Canada." Canadian Review of Sociology and Anthropology, 24 (4), 465-488.

Brady, Henry E. and Richard Johnston. 1987. "What's the Primary Message: Horse Race or Issue Journalism?" In Gary R. Orren and Nelson P. Polsby, eds., The New Hampshire Primary and Nomination Politics. Chatham, New Jersey: Chatham House.

Converse, Jean M. And Stanley Presser. 1986. Survey Questions: Handcrafting the

Standardized Questionnaire. Beverley Hills: Sage

Dillman, Don A. 1978. Mail and Telephone Surveys: The Total Design Method, New York: John Wiley and Sons.

Dunkelberg, William C. and George S. Day. 1973. "Nonresponse Bias and Callbacks in

Sample Surveys." Journal of Marketing Research, 10, 160-168.

Fitzgerald, Robert and Linda Fuller. 1982. "I Hear You Knocking But You Can't Come

In: The Effects of Reluctant Respondents and Refusers on Sample Survey Estimates." Sociological Methods and Research, 11 (1), 3-32.

Frey, James H. 1983. Survey Research by Telephone. Beverly Hills: Sage Publications.

Groves, Robert M. and Lars E. Lyberg. 1988. "An Overview of Nonresponse Issues in

Telephone Surveys." In Telephone Survey Methodology, Editors: Robert M. Groves, Paul P. Biemer, Lars E. Lyberg, James T. Massey, William L. Nicholls II and Joseph Waksberg. New York: John Wiley and Sons, 191-212.

Groves, Robert M. 1989. Survey Errors and Survey Costs. New York: John Wiley and Sons.

Hawkins, Darnell F. 1975. "Estimation of Nonresponse Bias." Sociological Methods and Research, 3 (4) 461-488.

Holbrook, Thomas M. 1996. Do Campaigns Matter? Thousand Oaks, California: Sage.

Johnston, Richard, André Blais, Elisabeth Gidengil, and Neil Nevitte. 1996. The

Challenge of Direct Democracy. McGill-Queen's University Press: Montreal.

Johnston, Richard, André Blais, Henry E. Brady, and Jean Crête. 1992. Letting the

People Decide: Dynamics of a Canadian Election. McGill-Queen's University Press: Montreal.

Kalton, Graham. 1983. Introduction to Survey Sampling. Newbury Park, California:

Sage.

Katosh, John P., and Michael Traugott. 1981. The consequences of validated and self-

reported voting measures." Public Opinion Quarterly, 57(1): 1-28.

Kish, Leslie. 1965. Survey Sampling. New York: John Wiley and Sons.

Lessler, Judith T. and William D. Kalsbeek. 1992. Nonsampling Errors in Surveys.

New York: John Wiley and Sons.

McDonald, P. 1979. "Noncontact and Refusal Rates in Consumer Telephone Surveys."

Journal of Marketing Research, 16, 478-484.

O'Rourke, D. and J. Blair. 1983. "Improving Random Respondent Selection in Telephone

Surveys." Journal of Marketing Research, 20, 428-432.

Pineo, P., J. Porter and H. McRoberts. 1985. "Revisions of the Pineo-Porter-McRoberts

Socio-economic Classification of Occupations for the 1981 Census", Program for Quantitative Studies in Economics and Population Report. No. 125. McMaster University.

Schuman, Howard and Stanley Presser. 1981. Questions and Answers in Attitude

Surveys. Orlando: Academic Press.

Tremblay, Victor. 1982. "Selection from Telephone Directories: Importance and

Consequences of Unlisted Numbers." New Surveys, 7 (1) 8-14.

Wiseman, Frederick and Marilyn Billington. 1984. "Comment on a Standard Definition of Response Rates." Journal of Marketing Research, 21, 336-338.

1. In addition to the surveys described in this documentation, the CES project included semi-structured in-depth interviews with a subset of the campaign-period respondents, interviews with party strategists, cataloging of election advertisements, as well as the collection of media coverage of the election.

2. ² Residents of old age homes, group homes, educational and penal institutions were excluded from the sample.

3. ³ Using their Household Inventory and Facilities and Equipment (HIFE) surveys, Statistics Canada estimates that two percent of the private households in Canada do not have a telephone (Ottawa, 1991).

4. ⁴ See O'Rourke and Blair, 1983; for a review of the birthday selection method.

5. ⁵ Weighting to correct for unequal probabilities of selection, stratification, and other factors in order to improve sample estimates is common in survey research. See, for example: Lessler and Kalsbeek, 1992 Chapter 8; Kalton, 1983 Chapter 10; and Babbie, 1992 Chapter 5. Kish, 1965; specifically addresses the issue of weighting to correct for unequal probability of selection at the household level (p. 400) and suggests, unlike most survey researchers, that household weighting may not be necessary.

6. ⁵ The household weights have been calculated using the household size information for the complete sample. Calculations of the household weight variable for Quebec only, or for Canada without Quebec, indicate that the household weight variable need not be recomputed for each sample component. The distribution of the population by household size is approximately the same in Quebec as it is in the other provinces and territories.

7. Respondents outside of Quebec were not asked how important "defending the interests of Quebec" was to them personally in the election.

8. ⁷ The 24 orders represent all possible order combinations for four items as determined by 4 factorial (4 x 3 x 2 x 1 = 24).

9. Respondents in provinces other than Quebec were not asked to rate Gilles Duceppe.