Measuring Survey Data « Back

How do you measure responses to a survey?

For most business requirements, where stakeholders reach conclusions and make decisions based on immediate survey data, descriptive measures help to summarize and describe data in a form that is manageable and assist in identifying patterns and trends. These measures typically include distribution (frequency, percentage), central tendency (median, mean) and dispersion (standard deviation).

The ability of these measures to perform will largely depend on advance planning of the survey structure for results that will meet the information requirements of stakeholders.

One approach is to measure change in responses over time (i.e., surveys taken on two different occasions).

Another approach is to compare responses between groups and within groups in the same survey. (i.e., survey taken by multiple respondent groups with questions in common and shared attributes).

A powerful variation is to measure difference between two sets of responses made at the same time about a set of issues or questions (i.e., two sets of answer columns for each question). The added value being a real-time measure through which business can assess change across a scale or continuum. Depending on the set-up of the question, this might be a change in opinion; an increase or decrease in skills; an improvement or deterioration in performance.

E-valuate-IT is a powerful tool that enables respondent groups to be asked different questions within the one survey. Where questions are the same, E-valuate-IT customized reports can be created comparing the groups' responses. Respondent attributes can be cross-tabulated within groups and between groups.

The Gap assessment by E-valuate-IT for double-column questions also provides cross-tabulation within groups and between groups. Comparison between questions is provided by the Gap assessment summary.

All customized and cross-tabulated summary reports can be exported. Survey response data can also be exported in a form that suits statistical packages.


Registered clients and partners,
LOGIN here

  Privacy Statement | Disclaimer