A somewhat old debate in psychology is whether or not the subject should be seen as a science. Wilhelm Wundt is considered by most to be the father of modern psychology, a scientific discipline, developed from the days of philosophical debates. With the incorporation of scientific techniques, the majority of psychologists see the subject as a science and follow the guidelines of scientific research.
The most important feature of scientific research is that it should be empirical, valid and reliable. To achieve this, an important factor to consider when conducting research is the types of bias and how to control for them.
- To start with, we will discuss the types of bias in research.
- Then, we will move on to discuss the importance of elimination of bias in qualitative research, quantitative research and data analysis.
- Throughout, we will look into research bias examples to help you learn and understand the types of bias and how to control for them in psychological research.
It's important that researchers consider potential issues with bias when conducting their studies, freepik.com/storyset
Types of bias and how to control for them in research
Bias can be a serious problem if not dealt with in research. If bias is found to be an issue in a study, then the study can be discredited for lacking credibility.
Bias in psychology research refers to when a researcher's subjective opinion influences research, such as their preferences or tendency for research to favour a sub-group.
Psychological research aims to follow the scientific method. According to the scientific method, three key components that make research scientific are that research should be empirical, valid and reliable. Bias in research can impair all three of these.This shows the importance of identifying sources of bias in research and identifying ways to control the bias.
Types of bias in research
There are different types of bias in research. Some types of bias are characterised by the researchers' actions, whereas other biases are apparent from how the study was carried out.
- Some of the types of bias in research that will be covered are:
- gender bias
- cultural bias
- age bias
- experimenter bias
- observer bias
- bias in questioning
Let's start by discussing the types of bias in research caused by the researcher's actions.
Experimenter bias is when the researcher sways the study to go in a certain direction because of their expectations of what the results should be. Experimenter bias affects how the experimenter behaves towards the participants and what results they deem appropriate, resulting in errors.
Experimenter bias lowers the validity of findings and suggests that the research may not be empirical. As the results are not based on what the researcher directly observed, and it is not an accurate representation of the relationship between the independent and dependent variables.
Another type of bias can be identified when observational studies are carried out. When analysing observational research, the researchers' subjective opinions can easily influence the results.
For instance, how the researcher interprets a participant's behaviour can be biased. If they are from different cultures or have different religious beliefs, or even if they interpret a behaviour incorrectly.
This is called observer bias.
Another type of bias in research that can affect the credibility of research is bias in questioning. How the investigator asks a question can lead to biased results. Misleading questions can cause participants to respond in a specific way; this reduces the validity of the findings.
Loftus and Palmer (1974) carried out a study to investigate how leading questions could affect the accuracy of participants' memory recall. The question changed when deciding what adjective to use when describing a car crash.
The research found that people were more likely to overestimate the speed of a car after watching a video clip of a car crash and asking how fast was the car travelling when the car 'smashed' each other compared to other adjectives.
However, the participants who were asked how fast was the car travelling when the cars 'contacted' each other estimated a lower speed. It is important to note that each participant was shown the exact same video clip.
From the Loftus and Palmer (1974) results, it can be inferred that people's responses can be influenced by how a question is worded. This shows that bias in questioning can affect the reliability and validity of the research.
If the wording of questions is different for participants, then participants' responses may differ, reducing the study's reliability.
Research is considered unbiased when it is based on empirical evidence that has not been influenced by the researchers' bias, freepik.com/rawpixel.com
Let's now move on to discuss the types of bias in research that are known to be a result of sampling errors.
The following types of bias are caused by researchers using a non-representative sample that is not generalisable to the broader population. These types of biases lower the generalisability and reliability of research. If these types of bias are apparent in research, then the findings may not be consistent and/or applicable to the sub-group that was not considered.
The different types of sampling bias are gender bias, age bias and culture bias.
Gender bias is when the researcher generalises findings based on one gender to another without empirical evidence.
Research carried out only on men is called androcentric, and these findings should not be generalised to women.
Vice versa, research carried out only on women (gynocentric) should not be generalised to men.
If a study is carried out only on young people, it can be considered to be age biased. These results would not be generalisable or representative of middle-aged or elderly people, and similarly, research carried out on an elderly sample cannot be 'generalised' to children.
Culture bias is when research is carried out in one country or only in certain cultures, and the results are then generalised to the entire population.
There are clear cultural differences in behaviour, mental health, and motivations, so research should consider these before generalising results.
For research to be established as unbiased than researchers need to recruit participants from various cultures, genders and ages, freepik.com/rawpixel.com
Elimination of bias in qualitative research
Consider the following table. Here, we discuss examples of types of bias and how to control for them in qualitative research.
| Type of bias | How to control for them |
| Observer bias | When analysing observations, the results should be compared by more than one observer. From this, the researcher can identify if there are differences between the results (which suggest low reliability). |
| Sampling biases | Recruit a sample representing males and females of all cultures and ages unless the research specifically investigates a sub-group of people. If this is not possible, results can only be generalised to the population the sample accurately reflects. For example, a sample of British children aged three to six will only represent British children in the results. |
| Bias in questions | Standardising methods, the same questions in the same order should be asked to each participant. The questions asked should also not be biased. |
Elimination of bias in quantitative research
Consider the following table. Here, we discuss examples of types of bias and how to control for them in quantitative research.
| Type of bias | How to control for them |
| Experimenter bias | Blind experiments, a common experiment type, is where a researcher compares finding from a control group and the experimental group. |
| Sampling biases | Recruit a sample representing males and females of all cultures and ages unless the research specifically investigates a sub-group of people. If this is not possible, results can only be generalised to the population the sample accurately reflects. For example, a sample of British children aged three to six will only represent British children in the results. |
| Bias in questions | Standardising methods, the same questions in the same order should be asked to each participant. The questions asked should also not be biased. |
Overall, both quantitative and qualitative research have similar ways to control for biases, as both are valid forms of research that approach data differently. Controlling for bias is viable for both forms of research.
Checking for alternative explanations, using blind experiments, verifying your research through past and potential future questions, and having others check the information are all good forms of controlling for different types of bias.
Research bias examples
Below is a hypothetical study; what type of biases do you think are present in the study?
A study was carried out on 23-25-year-old American men to investigate how childhood memories are recalled in later life. One of the questions asked in an interview was, do you recall any negative events from your childhood?
The types of bias present in this research are age bias as it was carried out only on people aged 23-25, this means the results founds may not be the same when tested on younger or older people.
In addition, to this the study can be considered to be culturally biased as the sample consisted only of Americans; there may be cultural differences found if the study recruited participants from various countries.
As the study was only carried out on men the study can be considered to be androcentric; this means that the study does not take into consideration gender differences that may occur.
Finally, in this research, there is evidence of bias in questioning as the researcher asked "do you recall any negative events...". From this question, participants may pick up that the researcher is trying to investigate negative events and so answer accordingly; this is problematic as it reduces the validity of the findings.
Types of bias and how to control - Key takeaways
- Bias in psychology research refers to when a researcher's subjective opinion influences research, such as their preferences or tendency for research to favour a sub-group.
- Bias can affect a study's reliability, validity and whether it can be considered empirical.
- There are different types of bias in research;
- sampling issues can cause age, cultural and genderbiases.
- experimenter bias, observer bias and bias in questions are other types of biases.
- There are different types of bias, and for each of them, there are different methods researchers can use to combat them. Blind experiments, random sampling, verification by other observers/experimenters, and using past and potential future research questions and results can all aid the control of different types of bias.
Explanations 
Exams 
Magazine 
