Consumers Opportunities of and Interest in Purchasing Green Electronic Products

Consumers Opportunities of and Interest in Purchasing Green Electronic Products

Bilag A. Methodology description of questionnaire survey

Gallup was in charge of realising the survey. This procedure ensured that the respondents were completely anonymous in relation to the Danish EPA.

Target group and selection of sample
The survey population consists of all people responsible for purchasing electronic products in households.

The survey is based on a sample totalling 1,000 respondents. The sample was selected from Gallup’s ongoing telephone omnibus survey. This procedure ensured that a representative sample of people responsible for purchasing participated in the survey. Subsequently, the respondents received a mailed questionnaire with a stamped envelope addressed to Gallup.

Design of survey and data processing
The survey results were compiled through a mailed questionnaire. This collection method was selected because it makes it possible to pose a wide range of comparatively complicated questions.

The data was compiled in the period 17 May to 20 June 2001.
The questionnaire was forwarded together with an accompanying letter in which Gallup informed the recipients of the background and purpose of the survey. Gallup also sent a reminder letter stating that the date of submission for the questionnaires had been extended by about a week.

Gallup designed the questionnaire jointly with CASA and the Danish EPA.

The questionnaire encompasses two types of questions:

Questions of facts - i.e. questions where respondents are asked to answer, e.g., “yes” or “no” to whether they have purchased an electronic product or similar.

Intensity questions - i.e. questions where respondent are asked to grade how important, for instance, a certain parameter was in the purchasing situation. For these questions, the survey typically used a four-degree scale from “very high importance” to “no importance”. In the analysis, the answers were cross-referenced so that all essential interrelations could be elucidated.

Survey validity and systematic errors

The percentages stated in the results section are subject to some uncertainty, because the survey only questioned a small sample of all people responsible for purchasing rather than all of them.
The uncertainty can be stated as a 95% margin of certainty. This means that the true percentages with 95% certainty are as follows:

Click to see html-version of: Tabel

From a high number of respondents, 700 randomly selected respondents participated in the survey. Out of the 700, 280 reply affirmatively to a specific question, i.e. 40%. As the table shows, the margin of certainty lies around 3.6%. This means that it is almost certain (95% probability) that the percentage of affirmative answers would lie in the interval between the two margins 36.4% (40%-3.6%) and 43.6% (40%+3.6%), if the entire group had been asked.

The analysis takes the statistical uncertainty into account, and therefore only mentions the statistically significant results. Besides the statistical uncertainty, there are two other sources of uncertainty in the survey, i.e. systematic errors and validity.

Systematic errors mean that groups of the population may systematically register a lower response percentage than other groups, thus figuring in the survey with lower weight than they should. For this survey, a total of 1,000 questionnaires were sent out. Of these, an estimated 960 questionnaires fell within the target area (the rest were returned as address unknown etc.). 709 replied adequately to the questionnaire, corresponding to a response rate of 74%. This response rate is satisfactory. No systematic errors were seen in the target group, and overall we can conclude that the results are representative of the population of people responsible for purchasing electronic products.

Validity addresses whether the study actually answers the specific questions asked. To ensure high validity in the survey, a steering committee set up for the purpose reviewed the questionnaire.

Inquiries received in the course of the survey showed that the respondents rarely ran into many misunderstandings or questions of doubt in the questionnaire.