Bias in data collection can be introduced by the researcher or other people involved in the data collection process. This can occur during sampling or during the construction of the instrument, as well as during the submission and gathering of data.
Data collection process is an important phase of research since all the data results, analysis, evaluations, and conclusions depend on the data collected. It is necessary that the data collected is free from bias and errors. The researcher should train the people involved in the data collection process in order to ensure that they gather bias and error free data. Bias can be of various types some of them are listed below.
Types of biases in data collection
Selection bias can occur during the selection of the population or the sample for the study. The population that is selected for the study should be most representable of the situation. The population should be reliable and every individual in the population should be willing to participate in the study.
Another form of selection bias can be when the researcher selects the sample from the population. Random sampling ensures that every element in the population has a fair chance to be selected. Non-random sampling may not represent the population completely. The researcher can select any element of the population but in this way it will not be representable of the population.
A volunteer bias can occur if the researcher only asks some volunteers to participate in the study. The researcher does not know that by doing so the sample will only be comprised of a particular group and will not be representable of the total population that is being studied. This can happen if a researcher takes a sample of volunteers that come to hospital to donate blood the researcher might think that it is easier to get the sample but the sample is not representable.
Instrument bias is also one of the most common bias in research. Instrument bias may occur due to lack of knowledge to construct the instrument. The way questions are directed to the respondents highly influence how they will answer. It is very important that the researcher only ask unbiased positive questions. Some questions lead to some answers that may result in biased answers. The questions should have only one meaning to the respondents. Language of the questions should be according to the understanding level of the respondents.
Study personnel bias
It is very important for the people involved in the administration and collection of data to have sufficient training and expertise. In case, there are many people involved in the data collection the researcher should make sure that he provides them training for all the process. A prior training will help them understand the procedure and their will be a uniformity in the way the personnel conduct the process.
Bias can occur when different people involved in the process of data collection use different standards. This shows that they do not have prior training and knowledge of how things should have been done. Bias can also occur if the study personnel have any conflicts with the respondents regarding opinions and beliefs. The study personnel should be informed that they should only record the true answers of the respondents. They should not taint the responses with their own opinions.
Respondents loss bias
A bias can occur in the research if any one or more respondents lose interest and do not retain throughout the study. The researcher should have a plan if a respondent leave during the study. The researcher should know the answers to these questions before hand: What alternative actions the researcher is going to take in order to fulfill the gap? What are the ways that the researcher can employ to better ensure that the respondents will retain throughout the study? In what ways the researcher can contact the respondents if they do not show up on the scheduled time?
In any case, if the respondent do not participate and leave the study, the study does not remain the same. The researcher can provide incentives to the respondents in order to ensure their participation. It also depends on the demographic characteristics of the population that how the researcher should keep the participants under control throughout the study.
Ways to reduce bias in data collection
There are many ways the researcher can control and eliminate bias in the data collection. The researcher should be well aware of the types of biases that can occur. This will help the researcher better understand how to eliminate them.
Objectivity is the key to avoid any bias in the data collection. The researcher should know the requirement of the research and should construct the instrument objectively. The questions should be formulated in a logical, clear, direct and positive wordings. Any external factors should be controlled that can impact the data collection process. If there are some external factors that could not be avoided as in the qualitative research, the researcher should weigh their impact on the data.
Use objective data source wherever possible. If subjective data has to be used corroborate it with other methods to improve the trust in data.
User-friendly research instrument
The research instrument can be a questionnaire, interview, experiments, or observation etc. The research instrument should have clear directions and easy to use for the respondents. This will ensure better responses from the respondents.
Train research personnel
Train the people who are involved in the data collection process. They should have knowledge of how to tackle difficult situations. They should know the type of language they should use with the respondents. They should also be willing to do the data collection for you, their willingness is important.
Avoid data entry errors
Be vigilant in recording the data sometimes the researcher miss some part of the data. The data that is missed cannot be fully recovered. It eventually results in bias in the data evaluation. You should have a complete plan how you are going to record data. If writing the responses in an interview is not feasible use a video recording or at least tape recording so that you do not miss any data.
Know your plan
The researcher and other people involved in the data collection should know the goals and objectives of research. They should also be aware of the importance of the transparency in data collection process.
- Simundić A. M. (2013). Bias in research. Biochemia medica, 23(1), 12–15. doi: 10.11613/BM.2013.003
- Pannucci, C. J., & Wilkins, E. G. (2010). Identifying and avoiding bias in research. Plastic and reconstructive surgery, 126(2), 619–625. doi:10.1097/PRS.0b013e3181de24bc
- Berk, R. A. (June, 1983). “An introduction to sample selection bias in sociological data”. American sociological review, 48(3), 386-398.