In academic research, if a college student desires to find out the musical preference of everyone studying in the university then it is possible. They can use non-probability samples, where the volunteers are easily available, so the survey is quick and expensive. It is a great option for pilot studies, qualitative research, case studies, and developing hypothesis for a future survey.
Here is a college students guide to market surveys and sampling online. Based on the above example, here are some ways to perform market research.
Upload the music lover’s survey questionnaire online asking pupils to vote for their preferred music type. The volunteer sample includes people participating in the survey on their own. This sample is guaranteed to be prejudiced and unfair because people with strong views about a problem search for ways to voice it. The values collected from such a sample are under or overstated and useless when you consider the big picture. The volunteered samples offer information about themselves, so this cannot be generalized to a large group.
Ask students personally to answer your questions associated with musical preference. It is also a biased sample because certain types of students are chosen. Even the location chosen impacts the result. For example, standing before Fine Arts building for a survey on musical preference will generate less response.
Randomly choose the sample and email musical preference survey questionnaire to those students. A sampling frame is a list of students to be sampled. If you took email addresses from your class then a bias will occur as it does not match the population of interest [every student in your university]. Have a sampling frame that aligns well with the population of interest as close as possible.
Systematic sample versus simple random sample
From the university’s directory obtain email addresses and email music poll questionnaires to every 40th name on the list. It is called systematic sampling. In SRS [simple random sample], people are selected randomly. There are no sibling students with the same surname selected in SRS but in systematic sampling, there are high chances of selection. Therefore there are subtle chances of bias in systematic sampling.
SRS and volunteer response
If you choose SRS plan to email music poll then as long as every respondent responds the sampling is not subjected to bias and can successfully represent the population of interest. In case only 40% of the respondent emailed back their vote then the poll results will not represent the population of interest because the issue is with volunteer response.
Consider a follow-up requesting politely for cooperation. It can possibly enhance response rate and the same sample can turn out to be a fair representation of the population of interest. The nonresponse is a problem but you managed to reduce the influence on the results.
For academic research use of online sampling helps for a real-time and quick survey. There is no need to depend on unreliable and slow DIY tools or email surveys but collect from experienced panel providers with online research panelists keen to support your study.