A group of interest is one that will be reported on separately (the Population size). For instance, it could be all family law advice clients, or all clients aged 65 or over. If you are sampling on a monthly basis, but will only report a full year's worth of findings, use the total number of clients in the year and refer to our Intoduction to Sampling for Client Surveys for assistance with monthly sample size calculation.
What percentage is expected to give the findings of most interest (the Sample proportion)? For example: "70% of clients strongly agree they would recommend the provider to others". Use previous survey findings if available. Otherwise, use 50% as the default, as this returns the largest sample size required.
The findings from survey samples are estimates of the true population percentage. What is the acceptable range around the percentage (the Confidence interval)? For example, a +/-5% range around 70% would mean a reasonable degree of confidence that the true percentage lies between 65% to 75%. We recommend +/-3% as a default.
How confident do you want to be that the true percentage in the population sampled is within the range selected above (the Confidence level)? The percentage selected here tells you that if you conducted 100 surveys, the number of times the true population percentage would be within the range selected above. Enter a number above 80%. We recommend 95% as a default.
As not everyone who is asked to participate in the survey will do so, the sample size must be increased to take account of this. Enter here the percentage of clients you expect to complete a survey (the Response rate). Use information from previous similar surveys, or if not available, use 20%.