A larger sample can yield better results. But, there is a cost involved for every respondent, in direct costs to conduct the survey and the time invested in their response. It is not always practical, or necessary, to ask everyone what they think.
If you choose to sample, a Sample Size Calculator can help. Here are a couple you might find useful. Each provides simple definitions for the terms used.
On top of what is “right” statistically in terms of sample size, you should also consider these other factors when determining sample size…
- Inclusion – Most of our clients are surveying employee groups, checking on their engagement or specific workplace issues. When you choose to sample employee opinions rather than asking everyone what they think, be concerned about the message is being sent to those not included. That doesn’t mean you need to include everyone. Just be careful that your reasons for sampling are valid, the criteria for selection can be explained to those who might ask, and include this information in communication to staff about the survey.
- Response Rate – Everyone you invite will not respond to the survey. The percentage of people who do is called your “response rate”. Estimate your response rate to determine how many invites to make in order to get the quantity of information you desire. Response rates vary for many reasons. We can help with some things, for example: length of the survey; quality and complexity of the questions; timing of the invitation; personalizing the request; and, well-timed reminders. Your relationship with the survey audience, both long-term and directly leading up to the survey, is also a key factor and mostly in your hands. Do what you can to ensure the audience trusts both your intentions for surveying, and your commitment to follow through with constructive action on what you hear.
- Cost – The greater the number of people surveyed, the higher the cost for conducting the survey typically. We charge a per response fee because more people responding means more work. But we try to help by only charging for responses, not invitations, and by keeping that response fee to a minimum.
- Purpose – It does cost more to survey more broadly. And statistical theory might say you’ll get a reliable response with the right sample. But you should ask yourself, “Why are we doing this survey – really?” Our clients are usually more concerned with hearing what the people surveyed have to say, rather than ‘proving’ anything statistically. The additional investment to seek out everyone’s opinion just may be worth it.