These consist in establishing the frequency distribution of each value (possible answer) for each variable (question). Potential mistakes can be detected on these frequency charts, the advantage of which is to draw a first picture of the results. They are generally delivered with graphs for a global overview of results.
These consist in determining the frequency distribution for all variables distinguishing them through the values of another variables, generally descriptive (age, gender, socio-professional group). E.g.: all results of the study crossed by gender of interviewees. As the name suggests, « non-responses » give the number of interviewees who did not answer a question, for each question.
These « non-responses » can be interpreted in various ways:
Frequency tables:
example
Cross tabulations:
These charts are generally delivered with the chi-square statistics, showing whether both questions are linked/related to each other or not. However, the chi-square does not show either the direction or the intensity of this relation.
example
“non-responses” interpretation:
It may be that people chose not to answer the question because they do not agree with any of the proposed values.
It may also be that people do not want to answer the question.
The « non-responses » must be as few as possible. Effectively, in any case, they will be the consequence of mistakes on behalf of the researcher.
Marketest offers different results analysis based on the clients requirements and budget. The results can be given in a basic way (individuals’ answers) or in a full report (frequency tables, graphs, cross tabulations and comments).
