There are many things that can influence a study along the way. Especially research bias is a topic that should be addressed in this case because warped results always make your study unreliable and in the worst case worthless. One type of bias is especially important, as it impacts the researcher as well as the participants. This type is called observer bias, which will be explained thoroughly in the following article.
Observer Bias – In a Nutshell
Observer bias describes the phenomenon that a person often sees what they want to see, and that these expectations can also influence the results of a study.
Definition: Observer bias
Observer bias is when observers tend to see not what is there, but what they want or expect. As a result, it leads to a systematic difference between the actual value and the observed one. This divergence typically stems from a researcher’s conscious or unconscious prejudices. When this type of research bias manifests, it can lead a scientific investigator to portray findings that stray from genuine accuracy and verity.
- ✓ Free express delivery
- ✓ Individual embossing
- ✓ Selection of high-quality bindings
Impact
The impact of observer bias can influence the researcher as well as the participants in their perception of a situation. This can lead to misinterpreted information and an inaccuracy in the description of reality. If the answers of the participants do not match the actual situation, the researcher will draw wrong conclusions that can affect the measurements taken based on the results.
Source
The source of observer bias is of course the humans themselves. Although it might be evident that observer bias will influence subjective methods, objective ones can also autumn victim to this type of research bias.
- Subjective research methods, especially in behavioural science, require of an interpretation to produce results, which means they are prone to bias. Every researcher has their own way of weighting events, which greatly influences the results of their observations.
Example
For example, if a group of researchers watches the behaviour of kittens in a new environment, they may focus on how their expectations were fulfilled by the small cats. This does, of course, not happen consciously, but people tend to notice things more when they expect them to happen. If they estimate the kittens’ ears to turn more to investigate their surroundings, the researchers might focus more on their ears than on the rest of their body language. It can also happen that one person interprets their behaviour as curios, while another one deems it as a sign of insecurity.
- Objective methods seem to have a less high probability of being biased. However, as there are still humans conducting the study, even objective studies can end up biased, if the researchers are not equally strict in measuring.
Example
When doctors take blood pressure readings, they have been found to round the reading up or down to a number of their choosing, which causes discrepancies in blood pressure measurement.
Prevent observer bias
Many types of research bias are impossible to avoid completely. This is also the case for observer bias. However, there are a few measurements that can be taken to minimize the risk of observer bias in your study.
- Blind or double-blind your study. This means that the researcher and the participants do not know who is in the treatment and control group. This way, the researcher’s expectations cannot influence biased treatment or their perception of the results.
- Use triangulation. This means that the researcher will use different methods to achieve certain information by exploring different datasets, methods, theories, etc. The more sources there are, the less likely it becomes that a study becomes biased.
- Multiple observers. One person can easily be influenced by bias without noticing. However, if more people observe the same scenario, each of them can compare their findings to the others, making them aware of different opinions and biases.
- Standardize procedures. If procedures are standardized and every researcher follows the same protocol or set of questions, the possibility of introducing bias is reduced.
- Random sampling. Sampling your participants randomly can also minimize the influence of observer bias because this helps blind the study.
- Specially training study observers. To further reduce observer bias, you can train the observers to be more aware of what influences their perception and what impact their prejudices have on the study. This can help them realise the bias while observing and staying more objective.
- Be diligent in your evaluation. After conducting the observation, make sure to double-check your notes and replay the whole scenario in your head. Sometimes you may be able to reflect on your own biased view.
Observer bias: Examples
Example 1: The case of the clever horse
One of the first recorded instances of observer bias was in the case of the horse, ‘Clever Hans.’ His owner, Wilhem, claimed that the horse could perform arithmetic equations. Wilhem would ask the horse questions, and Clever Hans would appear to answer by tapping its hoof with the correct answer.
A German psychologist and biologist, Oscar Pfungst, studied this phenomenon and found that when Hans neared the correct answer, Wilhem would subconsciously react in a certain way that would signal the horse to stop tapping. The expectations of the horse’s owner influenced the behaviour of Clever Hans and produced erroneous data.
Example 2: The case of the dull rats
In 1963, Kermit Fode and Robert Rosenthal conducted a study with senior psychology students who were each given rats to test. The researchers told the students that some rats were ‘bright’ and were bred to be good at solving mazes.
They also told them that the other rats were ‘dull’ and had been bred from rats that were particularly bad at solving mazes. The researchers then asked the students to run the rats through a maze and collect data. Over time, the ‘bright’ rats were twice as good at completing mazes correctly and quickly.
The researchers randomly assigned the ‘bright’ and ‘dull’ labels to the rats. The students’ preconceived notions about the rats’ abilities affected their interaction with them, which influenced their performance.
Types of observer bias
The types of observer bias include:
- The observer-expectancy effect is usually used in a research context to describe how the observer’s perceived expectations subconsciously influence the study participants’ behaviour.
- This effect may include influences on participants’ behaviour, including the creation of demand characteristics that affect the participants or selective recording of the research data.
- The observer-expectancy effect is also known as expectancy bias or the Pygmalion effect.
Example
Teachers in their classrooms may expect certain students to fail and others to pass. They may have formed these expectations through previous performances, observed behaviour, or comments from other teachers.
ㅤ
Unconsciously or consciously, the teacher may treat the students differently, which may lead to the students behaving and performing in a way that aligns with their teacher’s expectations due to observer bias.
- The actor-observer bias is a bias when one tends to attribute the causes of actions differently depending on whether they are an observer or an actor.
- As the actor, one’s behaviour is attributed to external factors, while as an observer, the behaviour of others is attributed to internal factors.
- It is an attributional bias that influences how we interact with and perceive others.
- This bias is mainly caused by the difficulty actors have in seeing their situation objectively, while as an observer, they can view a situation with perspective.
- It is also known as actor-observer asymmetry.
Example
Your class took an exam, and the results are out. When you realise you performed poorly, you blame external and situational factors such as a lack of a conducive study environment, lousy teaching, or bias from your teacher.
ㅤ
When you discover that one of your friends also performed poorly, you attribute their poor performance to laziness, truancy, or inattentiveness.
- The Hawthorne effect describes how a study’s participants alter their behaviour when they know they are being observed.
- This effect was discovered through research conducted at the Hawthorne Western Electric Plant (which is how it came to be named).
- The original research was initially designed to observe the effects of floor lighting on workers’ productivity. However, their productivity increased when the lighting was diminished and improved and when the researchers also changed other variables.
- They discovered that the increase in productivity was not because of a change in lighting but the increased attention from their supervisors.
Example
An example of the Hawthorne effect is a 1978 study that investigated whether cerebellar neurostimulators could reduce motor dysfunction in young patients with cerebral palsy. The results showed that the patients claimed a decrease in the dysfunction, even with the quantitative analysis showing very little increase in motor function.
ㅤ
They concluded that the increased interaction with the therapists, doctors, and nurses had had a positive psychological impact on the patients, which led to the illusion that they had tangible improvements.
Experimenter bias
Experimenter bias is a broad term used to cover the different types of biases that influence researchers in their studies. These biases include the actor-observer bias and the observer expectancy effect, among others.
FAQs
No, you cannot. It is impossible to avoid observer bias entirely, especially in studies where data is collected manually, but there are ways to minimize it as much as possible.
Researchers can minimize observer bias by standardized training of observers, using blinded protocols, identifying any potential conflicts of interest, and doing continued monitoring of objectivity in observers.
A researcher who has taken no steps to minimize this bias is more likely to misinterpret data. Observer bias has been shown to affect the validity of the results of studies significantly.
The Hawthorne effect describes a type of observer bias where the participants’ behaviour changes after realising they are being observed.
