Research Design

Questionnaire development

The starting point for the SPARC questionnaire was the questionnaire used in the American Cancer Society study.

The questionnaire has over 300 questions within these major sections: Attitudes and opinions, Your health, Health behaviour, Physical activity, Nutrition, Getting health and physical activity information, and About yourself. Changes were made by adapting the American questionnaire to suit the New Zealand audience.

The questionnaire was pre-tested with 22 respondents and the feedback from the pre-testing was incorporated into a final draft version for piloting. One hundred respondents were recruited in Auckland and Rotorua to pilot the survey independently. Around half the pilot questionnaires were received in time to improve the final questionnaire; 67 were returned in total.

Sampling

The random selection of 14,000 households was drawn from the electoral roll. Those of Maori descent identified in the electoral roll and addresses of those aged under 25 years were oversampled to try to counter the typically lower response from these groups.

Mailout (29 May-31 July 2003)

The mailout process involved a number of contacts with the selected households:

• A prenotification letter was sent out prior to the main questionnaire to inform the household of the pending survey.
• The questionnaire was then sent with a token incentive (pen) and a freepost envelope to send the completed questionnaire back.
• Approximately a week later, a 'thank you' postcard was sent to all households to thank those who had completed the survey and to remind those who had not yet completed and returned their questionnaire.
• A replacement questionnaire was sent three weeks later for those who had not returned the questionnaire.
• Three weeks later again, a final replacement questionnaire was couriered to households that had not yet returned a questionnaire.

Response rate

The final response rate of completed, usable questionnaires is 61%. Details are provided in the separate Technical Report.

The Technical Report also includes comparisons of the changing composition of responses (e.g. ethnicity) over the two-month survey period. The comparisons of response rates over the two-month period are unusually enlightening because of the large sample size.

Weighting

The final results are weighted by age, gender, and ethnicity to the New Zealand population. Details are contained in the separate Technical Report.

Analysis

The segmentation process was an iterative mixture of judgement and statistical analysis. Key processes included: selecting the target group; preliminary clustering and prioritising key variables to drive clustering; and splitting the target group into the final segments (details of clustering analysis are in the separate Technical Report).

The target group is a "middle" group including around 45% of respondents. It excludes those already regularly active for at least six months (45% of respondents) and those determinedly inactive (only active zero or one days and not even thinking about becoming regularly active; 9% of respondents). The analysis process is summarised in the research design diagram.

• View research design diagram.

Extra analysis to help identify main influences on behaviour was completed (using "path modelling"”). In particular, this was done to check the relevance to New Zealand of two possibilities suggested by the American Cancer Society research (Porter Novelli, 2002): that "intrinsic" types of motivation might be more important than "extrinsic" types; and that improvements in physical activity behaviour may tend to come at the expense of improvements with respect to diet (or vice versa).

Margins of error

The margin of error is around 2.3% for results about the full Target Group, and 5% to 7% for the six segments (of different sizes) into which it is split. Details are in Appendix II.

Conceptual overview

The diagram below summarises the major components (behaviour, personal factors, environmental factors) and fundamental interrelationships between them that underlie the differences between segments and that can be influenced to change levels of physical activity (following Maibach 1995 and Bandura 1998).

Conceptual overview diagram

Text description of the image

This diagram summarises the major components (behaviour, personal factors, environmental factors) and fundamental interrelationships between them that underlie the differences between segments and that can be influenced to change levels of physical activity.

Personal factors

The first major personal factor is perceived benefits. Perceived benefits include views on likely health benefits of physical activity (which are naturally affected by knowledge) as well as other possible outcomes such as looking better, having more energy, feeling more relaxed, and having fun.

Motivations are of two main types: intrinsic (e.g. I enjoy physical activity) and extrinsic (e.g. my family wants me to).

In addition, people's confidence about their ability to do a particular physical activity or regular amounts of physical activity ("self-efficacy") has been found to be important in changing behaviour.

If people are not confident that they can achieve an improved level of physical activity, why should they start trying? Confidence about making improvements can be changed by improving skills and strategies, and by setting appropriate goals.

Environmental factors

Environmental factors are mainly discussed in these reports as perceived barriers. These include not only barriers relating to the physical environment but also social barriers such as discouragement from others.

Note also all the two-way influences between behaviour, personal factors, and environmental factors. It may be obvious to think of improving physical activity behaviour by changing personal factors (e.g. beliefs about health effects) or environmental factors (e.g. improving facilities).

However, one should not overlook the effects in the other direction (e.g. effects of behaviour on perceived benefits).

For example, someone who goes for a long walk for the first time after lengthy inactivity may be reminded of how good it feels to
be active.

Contents | → Next: Focusing on Major Influences

Updated | 13 Jan 2009.

Skip page tools.

• Go to top