So far in the Research Methods topic, you've learnt about formulating hypotheses, variables and sampling methods. Now it's time to learn about designing research and the research methods available to do so.
Why is this important?
When psychologists plan their research, they need to consider research methods, namely, which method is best for their research. This can depend on many things; the hypothesis, the research subjects, the environment and so on. Due to this, there is no 'one size fits all' with regard to research methods.
- This topic will cover and summarise designing research; including the different research methods that are commonly used in psychology.
- We will start off by looking at the research design psychology definition and the purposes of research designs.
- Moving along to learn about the types of research design, these will include both quantitative and qualitative research designs.
- Throughout, you will notice research design examples
Various research methods can be used to investigate psychological phenomena, freepik.com/pch.vector
Research Design Psychology Definition
Research design refers to the research plan and procedure that is followed during a study.
The research design is a specific plan the researcher(s) uses for the:
- collection
- analysis, and
- interpretation of data.
Purpose of Research Design
What is the purpose of research design? The short answer to this question is that it helps researchers find answers to their research questions in the 'best' way possible.
Imagine that you are trying to find out about your class's preferences about ice cream. You could go about this by asking each person individually, but with a group of potentially 25-30 people, this could take time. You may not get straightforward answers. It may be better to consider creating a questionnaire or a poll.
Consider this on a large scale. With complex hypotheses, lots of
variables, and many research subjects, researchers have to be careful about how they carry out their research, as this can affect what kind of data they get.
Researchers have to plan the 'best' way to get the answers to their research questions. This is called research design. Pixabay.com
Quantitative and Qualitative Research Design
Research can be carried out in a multitude of ways, but how is it presented? Research can widely be categorised into quantitative and qualitative research methods.
Quantitative research methods are methods that generate numerical data.
Examples include questionnaires (through 'yes' or 'no' questions) and structured interviews.
Questions with open-ended responses such as how do you feel produce qualitative data. So keep in mind that not all questionnaires are quantitative methods.
Qualitative research methods are methods that generate mainly detailed, worded data.
Examples include unstructured interviews and case studies. Now, let's go over some research methods.
The experimental method
The experimental method in psychology refers to establishing cause and effect relationships using independent and dependent variables. This is done through the use of experiments.
Experiments, as you may already know, are tests of a hypothesis. Depending on the experiment's outcome, a hypothesis may be supported, disproved or need further testing.
Types of experimental design
When using the experimental method, researchers have to consider the experimental design. Experimental design refers to the method of allocation of participants in an experiment. There are three types of experimental design:
independent groups
repeated measures
matched pairs
We will briefly go through each of these.
Below you will get an overview of the types of experimental design. We won't cover their advantages and disadvantages here - that's for our 'The Experimental Method' article!
Independent groups design
The independent group design is an experimental design where there are two sets of participants that take part in different conditions of the experiment.
An experiment is conducted using two different conditions: the experimental group and the control group.
Group 1 (experimental) consists of Felix, Abby, Sana, Lucy and Abdul. Group 2 (control) consists of Luke, Andy, Julia, Amina and Enis.
In this experiment, participants were told they would eat a sugary tictac to see if it made them hyper. The experimental actually received the sugary sweet, whereas the control group received a 'fake tictac' that did not contain high sugar.
Repeated measures design
The repeated measures design is an experimental design where the same set of participants takes part in all the conditions of the experiment. The set of participants remains the same, even if experimental conditions are adjusted.
In an experiment about memory recall, Felix, Abby, Sana, Lucy and Abdul take a memory test with only picture clues to help.
They then take another memory test, this time without any picture clues.
Matched pairs design
The matched pairs design is an experimental design where participants are matched into 'pairs' due to key shared features, such as age, sex or biological characteristics.
Once these pairs are created, they must be randomly allocated to either the experimental group or the control group. Due to the key shared features between the paired participants, researchers will be able to compare the participants in different conditions.
In an experiment to test and compare the effects of two new medical drugs, a researcher chooses matched pairs design. Amina and Lucy are paired together because they share the same sex, age, height and weight.
Randomly, Amina is allocated to the experimental group (she takes drug A), and Lucy is assigned to the control group (she is given drug B). The researcher will then compare the effects and results of drugs A and B of the matched pairs.
Types of Experiments
It is also important, of course, for researchers to consider which type of experiment they will carry out. There are three types of experiments:
- laboratory experiments
- field experiments
- natural experiments
They differ in various ways; however, they have some common features. We will consider each type of experiment.
Laboratory experiments
Laboratory (lab) experiments are carried out in controlled, artificial settings. Out of all the different types of experiments, lab experiments are the most controlled.
In lab experiments, all extraneous variables are controlled by either removing them or keeping them in a certain state. In this way, researchers can clearly establish causal relationships by looking at the relationship between the independent and dependent variables.
Milgram's shock experiment (1963) is a famous (or perhaps infamous) example of a psychological lab experiment. As all of the extraneous variables were controlled, Milgram could easily establish a relationship between an authoritative figure (the independent variable) and the levels of obedience (the dependent variable).
Lab experiments have their advantages and disadvantages. One advantage is that lab experiments are very good at establishing cause and effect relationships (as mentioned above). They are also replicable as the conditions are controlled.
A disadvantage, however, is that they lack ecological validity due to their artificial settings.
Field experiments
Field experiments are experiments that are carried out in a natural setting. This means that the setting is not controlled. Examples of natural settings could be in a school classroom, on a train or in a dentist's waiting room.
Field experiments work by testing the effect(s) of an independent (manipulated or controlled) variable on a dependent (measured) variable.
An example of a field experiment is:
Researchers want to measure public attitudes to homeless people. They may set up an experiment where a (fake) homeless person enters a public space (such as a train carriage) and asks for money. The homeless person is the independent variable, and people's reactions (namely, whether they give money to or ignore the homeless person) is the dependent variable.
Field experiments have their advantages and disadvantages. One advantage is that they are high in ecological validity.
Contrast this point with laboratory experiments, which have low ecological validity.
One disadvantage is that not all extraneous variables can be controlled, and therefore it may be hard to establish causation between independent and dependent variables.
Natural experiments
Natural experiments are experiments that are also carried out in a natural setting. However, they differ from field experiments in one key way: in natural experiments, the researchers do not control an independent variable, so they must simply measure the dependent variable. This is because the independent variable already exists.
Researchers will measure the dependent variable and assume that this is due to the independent variable.
An example of a natural experiment is:
Researchers want to compare the brains of those suffering from alcoholism and those without alcoholism. In this case, they cannot manipulate the independent variable as it already exists, namely, the people suffering from alcoholism. Researchers will not be able to control who belongs in the 'alcoholic' and 'non-alcoholic' categories.
Field experiments have their advantages and disadvantages. One advantage is that they are also high in ecological validity as they are conducted in natural settings.
One disadvantage is that there may be a sampling bias as the independent variable cannot be manipulated.
Interviews
Interviews are an example of a qualitative research method. Psychological research often involves some interaction between the participant(s) and the researcher. This interaction could be:
- in person
- online
- over the telephone
Interviews are a good way of obtaining data for research purposes. Unsplash.com
Types of interviews
Not all types of interviews are the same; researchers have to consider what is best for their research and purpose. Does the researcher have questions prepared before the interview? Or will the interview be free-flowing?
Let's consider some different types of interviews.
Structured interviews
Structured interviews consist of close-ended questions; neither the participant nor the researcher can add to the question or answer, as the answer is designed to be concise.
The answer to the question "What is your favourite ice-cream flavour?" would be something straightforward and simple in a structured interview, e.g. "strawberry". You would not have the opportunity to elaborate further on this answer.
A researcher is likely to have a set of standard questions that they go through. For this reason, structured interviews may be carried out fairly quickly compared to other types of interviews.
Unstructured interviews
Unstructured interviews consist of open-ended questions; since the questions are not prepared beforehand, the interview can take the form of an informal conversation between the researcher and participant.
A question such as "How do you feel about the subject of History at school?" is a good question for an unstructured interview because it allows the participant to answer openly.
Depending on the participant's answer, the researcher may ask follow-up questions, such as "Why do you feel this way?" or "Can you elaborate on that?"
Semi-structured interviews
As demonstrated by the name, semi-structured interviews consist of a mix of structured and unstructured interview questions. The researcher may have a set of prepared questions, but they can also ask follow-up questions if deemed necessary.
Focus groups
Focus groups consist of a group of participants that are interviewed at the same time by a researcher. They can often take the form of a semi-structured interview.
Focus groups are popularly used by businesses to gauge the quality of their goods or services. They may ask a small group of people to, for example, test their new food product and conduct a focus group to get an idea about the quality of the product. They may ask questions about taste, texture, packaging, etc.
Questionnaires
Questionnaires are a popular method used in psychological research as they are a good way to collect data in a fast and inexpensive way. They can be administered:
- in person
- over the telephone
- using a paper questionnaire
- online
Questionnaires can consist of closed or open-ended questions.
Closed questions
Closed questions only accept answers from a set amount of options. Multiple-choice questions are a good example of closed questions, as participants must select from a fixed set of answers.
Open questions
Open questions allow the participant to add more detail to their answers. This may involve, for example, the questionnaire asking, "Please explain why you selected this answer". This may give more insight to researchers.
Questionnaires can consist of open and closed questions. Pixabay.com
Case studies
Case studies are another example of a qualitative research method. A case study is a type of research design that places a heavy focus on a single person, event, phenomenon or issue. The idea is to gain in-depth detail so that researchers can understand how the person or situation works.
Case studies may be suitable when researchers want to explore a new area of research or a new issue in more detail. They may also be suitable when researchers are trying to decide whether further research is needed in the area.
In psychology, a famous case study is the case of Little Hans conducted by Sigmund Freud. Freud analysed Hans' phobias through Hans' dreams and concluded that Hans was scared of his father.
Case study methodology
Due to the level of detail needed for case studies, case study methodology can be varied. Researchers can use triangulation, which is the use of multiple research methods. This can help researchers collect not only a wide variety of data but different types of data.
To uncover the reason(s) behind a particular event, researchers may conduct unstructured interviews to gain people's opinions as well as numerical data to understand how the event happened.
This can give researchers a mix of qualitative and quantitative data.
Observation Studies
Observation studies are often carried out in psychology to observe behaviours in various settings. Although this may sound like a simple form of research, there are many different types of observations.
Let's consider these in turn.
Participant and non-participant observation
Will the researcher be involved in the observation?
Participant
This is where the researcher themselves are involved in the activity that the participants are being observed in.
For example, if participants are playing a game, the researcher may decide to be the judge and observe the 'players'.
Non-participant
This is where the researcher does not take part in the activity and simply watches the participants.
Covert and overt observation
When designing observational research, researchers may consider whether the participants need to know that they are being observed or if it is better to remain 'undercover'?
Covert
This is where the researcher does not tell the participants they are being observed. Researchers could remain behind a one-sided mirror glass so that they can see the participants but not vice versa.
Overt
This is where the participants know that they are being observed.
Naturalistic and structured observation
In what kind of setting should the participants be observed?
Naturalistic
This is where participants are observed in a natural setting, such as in a park, classroom or at home. The researcher does not stage or change the setting in any way.
Structured
Structured observations are where participants are observed in a staged or altered setting in order to test the effects on behaviour. For example, a researcher may choose to observe children's behaviour when their parents are in the room versus when they leave the room.
Factors to consider when conducting observation studies
Researchers should consider categories of behaviour and inter-observer reliability.
Categories of behaviour
In an observation of many participants, there is likely to be a lot of activity and behaviour. How will researchers decide which behaviours to observe and which are relevant for the research?
Due to this issue, researchers should operationalise behaviours by creating categories of behaviour. Within these categories, researchers can look out for specific behaviours.
In a classroom observation, researchers create the following categories of behaviour for students working on group projects. They look for specific behaviours within these categories:
- Aggression (shouting, hitting, kicking, pushing, arguing)
- Collaboration (sharing resources, asking each other questions, laughing and/or chatting)
- Withdrawal (silence, sitting in isolation)
In this way, researchers can find and measure behaviours that are directly relevant to their research purpose.
Inter-observer reliability
What if one behaviour is interpreted differently by two researchers? Naturally, this would call into question as to which interpretation to follow for research purposes.
Therefore, when conducting observations, researchers should establish inter-observer reliability.
Inter-observer reliability refers to the degree to which two (or more) researchers agree on the behaviour that is being observed and recorded.
Inter-observer reliability can also be referred to as inter-rater reliability.
If two researchers are observing a classroom to measure levels of aggression and conflict between students, they may have different views on what 'counts' as aggressive behaviours. One may consider shouting as aggressive, and the other may not.
If they disagree with the observed behaviours, it is unlikely that their observation will be reliable.
To avoid this issue, it is important that researchers operationalise behaviours (as mentioned above so that they are both on the 'same page' in terms of what to look out for. If they do this, they will have inter-observer reliability, and both of their data sets will be reliable).
To continue from the above example, if researchers operationalise 'shouting', they will both be able to count how many times students shouted. They will have the same figure.
Designing Research - Key Takeaways
- The purpose of research design is to help researchers find answers to their questions in the 'best' way possible.
- Designing research involves considering both quantitative and qualitative research methods.
- The experimental method refers to the testing of the relationship between an independent and dependent variable. The three types of experimental design are independent groups, repeated measures and matched-pair design.
- The types of experiments are laboratory, field and natural experiments.
- Other research methods include interviews, questionnaires, case studies and observational studies.
Explanations 
Exams 
Magazine 
