5 Reasons You Are Not Getting The Data You Need

5-reasons-why-you-are-not-getting-the-data-you-need-1-768x512

Designing a research project takes time, skill, and knowledge. If you don’t approach the process with a clear goal and methods, you’ll likely end up with skewed data or a wrong picture of what you were trying to achieve. At Philomath Research, we simplify your data collection process.

While it is important to use proper methodology in the research process, it is equally important to avoid critical mistakes that can produce inaccurate results. In this article, we’ll list 5 common errors in the data collection process and show you how to avoid them, so you can get the best possible data.

 1. Population Specification

Population specification errors occur when the researcher does not understand who they should be surveying. This can be difficult because there are many people who may consume the product, but only one who buys it, or they may miss a section when looking for a purchase in the future.

Example: Packaged goods manufacturers often conduct surveys of housewives, as they are easy to contact, and it is assumed that they decide what is to be purchased and make the actual purchase. This situation often results in a population specification error.

The spouse may purchase a significant portion of the packaged goods and has a significant direct and indirect effect on what is purchased. For this reason, excluding husbands from samples may result in results targeted to the wrong audience.

How to avoid it:

Understand who buys your product and why they buy it. It’s important to survey the purchasing decision-makers so that you know how to best reach them.

 2. Sampling and Sampling Frame Errors

Survey sampling and sampling frame errors occur when the wrong subpopulation is used to select a sample, or because of variation in the number or representation of responding samples, but the resulting sample population is not representative of concern. Unfortunately, some element of sampling error is unavoidable, but sometimes, it can be predicted.

Example: Suppose we collected a random sample of 500 people from the general US adult population to assess their entertainment preferences. Then, on analysis, it was found to be made up of 70% of women. This sample will not be representative of the general adult population and will affect the data. Women’s entertainment preferences will bear more weight, preventing accurate extrapolation of the US general adult population. Sampling error is affected by the homogeneity that is being studied and the sample being sampled from and by the size of the sample.

How to avoid it:

While it can’t be avoided completely, you should have several people review your sample for an accurate representation of your target population. You can also increase your sample size so that you get more survey participants.

3. Selection

Selection error is the sampling error for a sample selected by a non-probability method. When respondents choose to participate in a study themselves and only those who are interested respond, you may end up with selection error because there may already be an inherent bias.

This can also happen when respondents who are not relevant to the study participate, or when there is a bias in the way participants are placed into groups.

Example: Interviewers conducting the Mall Intercept study have a natural tendency to select respondents who are most approachable and agreeable when they have the latitude to do so. Such samples often include friends and allies who share some degree of similarity in the characteristics of the desired population.

How to avoid it:

Selection error can be controlled by making extra effort to get participation. A typical survey process involves initiating a pre-survey contact requesting cooperation, an actual survey, and a post-survey follow-up. If no response is received, a second survey request follows, and perhaps an interview using alternative methods such as telephone or person-to-person.

 4. Non-responsive

Non-response error can exist when an obtained sample differs from the original selected sample.

This may be because either the potential defendant was not contacted, or they refused to respond. The main factor is the absence of data rather than incorrect data.

Example: In telephone surveys, some respondents are unreachable because they are not at home for the initial call or call-back. Others have moved or are away from home for the duration of the survey. Respondents who do not live at home usually do not have young children, and the proportion of wives working with someone at home is much higher. The geographic mobility of those who have moved or are far away for the survey period is higher than that of the average population. Thus, most surveys can infer errors from the non-contact of the respondents. Online surveys try to avoid this error through e-mail delivery, thus eliminating respondents not at home.

How to avoid it:

When collecting responses, make sure your original respondents are participating, and use follow-up surveys and alternative ways of reaching them if they don’t initially respond. You can also use different channels like in person, web survey or SMS to reach your audience.

5. Measurement

Measurement error is generated by the measurement process itself and represents the difference between the information generated and the information desired by the researcher. In general, there is always some small level of measurement error due to uncontrolled factors.

Example: A retail store wants to gauge the response of customers to a purchase over the counter. The survey has been developed but fails to target shoppers in the store. Instead, the results are skewed by customers who purchased items online.

How to avoid it:

Double-check all measurements for accuracy and make sure your supervisor and measurer are well trained and understand the parameters of the experiment.

While not all these errors can be completely avoided, recognizing them is half the battle. The next time you’re starting a research project, use this blog as a checklist to make sure you’re doing everything possible to avoid these common mistakes.

Also, before starting your next research project, make sure you take help and advice from a expertise of market research. This is important for any research project because you cannot start creating surveys until you understand the research problem.

Ready to start your research journey? Visit www.philomathresearch.com for further details.

Comments are closed