sequential analysis

KGS Cleans and Analyses data for accurate results

Sequential Analysis

This is a powerful yet simple tool that provides an analysis of cumulative results as they are obtained.

Provided questions are framed correctly sequential analysis can yield statistically significant answers with relatively small samples (8-25 respondents).

As well as helping solidify survey results where budgets (and/or samples) are small, sequential analysis allows intelligent decisions within ongoing studies by eliminating or further opening particular options available to respondents.

Sequential analysis can be applied to questions that have answers framed as a dichotomy: A or B, Yes or No (no preference/ don’t know replies are also possible).

Answers are plotted within boundaries that when reached will provide a highly definitive answer.

Consider the example here with Yes/No/Don’t answers recorded for 30 respondents.

‘Yes’ answers direct the blue plot toward the upper boundary and ‘no’ answers toward the lower boundary.

If either boundary is reached that answer is definitive (here with 95% confidence).

The plot shown doesn’t quite reach the upper boundary – however by response 25 the central boundary is a factor.

The blue plot cannot reach the lower boundary without crossing the central boundary.

The researcher can be 95% confident that ‘No’ cannot win and this choice can be eliminated after just 20 responses.