Calculation of a percent figure’s “true value”

Here you can test the statistical uncertainty of two percent figures from the same survey.


First percent figure you want to test
%
Second percent figure you want to test
%
Number of interviews you have in your survey
interviews
Number of interviews you have in your survey
interviews
The statistical certainty you want
%

RESULT:

The true value of your X% is between X% and X%
The true value of your X% is between X% and X%

Likelihood of a percent figure being different compared to another (same sample)

Here you can test how likely it is that two percent figures from the same survey are different.


Percent figure you want to test
%
Percent figure you want to test against
%
Number of interviews you have in your survey
interviews

RESULT:

The likelihood that your X% is actually different from your X%
is approximately: X% (using a 95% confidence level)

Likelihood of a percent figure being different compared to another (two samples)

Here you can test whether two percent figures from two different samples for example two different weeks or waves in a tracking are different.


Percent figure from “wave 1” you want to test
%
Percent figure from “wave 2” you want to test
%
Number of interviews you have in your “wave 1”
interviews
Number of interviews you have in your “wave 2 “
interviews
The statistical certainty you want
%

RESULT:

The difference must be at least X in order to be significant at X% confidence level
The difference is in fact: X
Thus the difference between the X% and the X% is !

Weekly Tracking Sample Size Calculation

Here you can estimate how many interviews you need in a weekly tracking, if you know how much difference from week to week you are looking for.
For example: if you expect your brand awareness to be 50% during week 1 and you want to be able to measure an increase to say 60% in week 2, and you want to measure this difference with 95% certainty, you will need at least 188 weekly interviews. Here you can test whether two percent figures from two different samples for example two different weeks or waves in a tracking are different.


The percent figure of a given answer in week 1
%
The percent figure of a given answer in week 2
%
The statistical certainty you want
%

RESULT:

If you want to measure a difference from X% to X%
from one week to another: you will need at least X weekly interviews.