Why Must We Do Research In Psychology?
1. Why must we do research in psychology?
Research in psychology is essential, as it consists of certain guidelines, methods and psychometric measures to ensure the findings produced are as scientific and reliable as possible.
Where certain social and psychological phenomena are commonly agreed upon, social scientists are able to challenge them using a scientific research method, producing factual findings that aim to be objective, replicable, free of bias and can be generalised.
Research methods used in psychology allow for obtaining quantifiable data, that can be processed, analysed and replicated. This enables researchers to generalise their findings and make predictions as to how and why certain behaviours occur.
If there was no research in psychology, certain theoretical concepts could not be measured, scientifically presented and the relation between them explained giving the most accurate and supportive evidence. Therefore, often abstract ideas need to be operationalised and tested in various settings using established psychological research methods.
2. Explain why we need to use statistics in psychology research.
In psychology, researchers are usually confronted with a very large sample of participants, which creates large amounts of data. If the large amounts of data gathered is unprocessed; it does not provide a clear representation of what the results of the study are and what they mean.
Information gathered, whether qualitative or quantitative, needs to be processed (i.e. mean, median, mode, standard deviation etc.), in order to achieve an arranged set of data, that would be easy to present, process and understand. A data set arranged in this fashion allows for noticing the differences and trends between groups, as well as subjects and establishing the significance of the findings.
Processed data allows for presentation in a graphic form (i.e. line graphs, pie charts, bar graphs, etc.), which allows the researcher to better overhaul the data, notice tendencies and patterns across the tested sample and describe them. The use of statistics in a graphic form presents the information in a “digested” way, making the output easy to understand for all parties even those out with the field of psychology (i.e. by going to a medical board with narrative in a niche area may not achieve the same results as going with statistics – a universal language).
3. Outline the key characteristics of psychology experiments.
An experiment is a type of study where, if carried out properly, the relation between cause and effect is measured.
The experiment is characterised by a controlled environment in which the experimenter can manipulate the independent variable.
In order to carry out an experiment one needs an experimental hypothesis, which would predict what the relation between independent and dependent variable is.
In experiments confounding and extraneous variables are ought to be controlled.
Experiments can also be characterised by the environment in which they are carried out. i.e. lab, field and natural.
In order to carry out a reliable and valid experiment one should ensure that control group is also used in the study.
4. What is ‘internal reliability’? [2 marks]
Internal reliability is a term that describes the consistency of a certain measure within the study. The internal reliability determines whether the method used in the study is consistently measuring the same thing and giving similar results within the tested group.
5. What is the relationship between a dependent variable and an Independent variable? Define each in your own words.
The independent variable (IV) is the one manipulated by the scientist and the dependent variable (DV) is the result of the manipulation that’s measured by the scientist. Both IV and DV are measured and their values can change, however, the difference in IV value is controlled by the researcher, whereas the DV value can only change as a result of the manipulation of IV.
6. Using an example of your own, what is a within subjects’ design?
Within subjects’ design is a type of experimental design where all participants are exposed to all levels of independent variable (IV). A good example for this would be the study discussed in class, where the experimenter wanted to investigate how driving skills are affected in different situations and used the within subjects’ design where the researcher exposed the same group to three levels of IV.
7. Under what conditions is it acceptable to use deception in a psychology study?
One of the key ethical guidelines in psychological research is to keep your participants free of deception. However, it is highly likely that if participants are fully informed, they might predict the methods used in the study and its real objectives and not give an honest and objective answer (social desirability bias, the ‘screw you’ effect). This puts a researcher in an ethical dilemma of what information should be revealed and to what extent.
Deception is acceptable, where the research study requires it to ensure unbiased and reliable data is produced. The participants, however, must be fully debriefed afterwards and reminded about the confidentiality, anonymity, privacy and the right of withdrawal.
Deception in research can be used as a method of prevention of producing artificial behaviours. In a psychological study, the use of deception is only justifed when obtaining unbiased information. This can be endangered by fully disclosing the true aim of the study. Only when other methods are not producing viable results, in terms of an unbiased outcome, then one should look to the use of deception.
The participants feelings and emotions should be taken into account when considering using deception. If revealing the true aim of the study during the debriefing causes mental or physical distress and discomfort to the participant, the use of deception should be reconsidered and possibly aborted.
Whenever the use of deception is considered, one should always consult with fellow researchers to ensure the use of deception is justifiable and that the trustworthiness of the study is not affected.
8. What are participant and experimenter effects and what can we do to minimise their effects on the results of our study?
There are a few participant and experimenter effects that can affect the reliability and validity of the study.
Demand characteristics suggests that participants are likely to adapt their behaviours and responses to what the aims of the study may reveal when interpreted.
Social desirability bias occurs when the participant decides not to reveal their honest responses and as a result answers in accordance to socially accepted behaviours.
The “yea saying” bias occurs when the participant shows the tendency to agree with everything regardless of the nature question or task given. These types of bias are best avoided by using single or double blinded method, ensuring both participant and researcher are unfamiliar with the research aims.
As mentioned above, deception allows to omit the social desirability or “yea saying” bias as it allows the researcher to keep the participants naïve and ensure as honest answers as possible.
The experimenter effects include confirmation bias which is based on the idea that the researcher is prone to analyse the results in favour of the hypothesis stated. The researcher can sometimes reveal the actual aim of the study with certain body language signs, tone of voice or wording of questions.
These types of bias can be controlled again by the double blinded method, minimising the contact between the researcher and the participant and/or getting the hypothesis and results re-evaluated by fellow researchers.
9. Psychology Researchers have to be realistic about the practical constraints and compromises that need to be taken into account when designing and carrying out research. Give examples of the types of factors that we need to consider.
There are a lot of constraints when conducting psychological research. Research in psychology requires participants and this creates the problem of finding a adequate sample size. Researchers also need to ensure that the sample is representative of general public in order to improve the quality and validity of findings, however, there is usually a problem of finding participants across all social strata, that are randomized and not opportunity sampled.
In this instance, the researcher has to sometimes compromise on how representative the sample is and mention that in the discussion to allow for further research in different (cultural, ethnic, age group etc.) settings in order to improve the sample and therefore reliability and validity.
The experiments conducted in a laboratory setting usually provide reliable results, however, the ecological validity of such a study is compromised. Taking the experiment into the field or natural setting improves the ecological validity but the reliability suffers. In this case the researcher has to compromise on one or the other and replicate the study in different settings to ensure that both reliability and validity are not compromised.
Research in psychology requires a large sample size, designing an appropriate method and the use of several tools to measure the outcome. The resources necessary for psychological research are usually expensive and require funding. In this case, the researcher needs to refer to larger companies and organisations for funding. This creates a problem of meeting your own agenda, ethics and objectives proposed by the organisation that provides the funding. In this case the researcher is faced with the dilemma of whether they should cut down on the resources necessary for the research and reducing the quality of the research or face the problems of expectations from the organisation that is funding the research and refer to further research that would ensure objectivity of the findings.
The studies in psychology are often longitudinal, where the researcher measures the change over several months and often years. This method produces great, detailed data. However, this method requires a lot of the researcher’s time, is expensive and often cannot be administered to larger group of people, since it can become unmanageable. In this case, the researcher can compromise and follow up only for a certain amount of time, describe the findings to date and highlight the fact that it requires further, longitudinal research.
In the above-mentioned type of study, the researcher also has to compromise on the reliability of the findings. As usually, the longitudinal study would produce rich, qualitative data that will apply to certain individuals but it is highly possible that the findings will not apply to general public. This is why further research has to be carried out on participants from various backgrounds.
10. What do we mean by the term ‘operationalising’ a variable?
In psychology, term operationalisation refers to the process of creating a method of interpreting the abstract ideas and putting them in an empirical, measurable and psychometric form (i.e. in order to measure one’s empathy). The theory of empathy cannot be measured directly as it is an abstract construct.
In this case, a researcher may operationalise this variable into a set of questions about one’s attitudes towards other people’s suffering, happiness etc. that carry certain weight depending on the answer chosen [1(strongly disagree) to 5 (strongly agree)].
11. What is ‘external validity’?
External validity is the extent to which a study’s findings can be applied to the general public. There are two main factors affecting the external validity, the setting in which the study was conducted and the sample used. The case for experiments shows how important external validity is. The laboratory setting experiments have been criticised as they can create a certain behavioural change in participants and not be reflective of the behaviours in general public (i.e. the well documented criticism of Milgram 1963).
12. Illustrate the key differences between experimental designs and
correlational designs using your own examples.
There are many differences between experimental and correlational methods. The main difference being that experiments measure the cause and effect, whereas correlation studies look at the relationship between x and y.
For example, in an experiment, the researcher creates the hypothesis, manipulates the IV and measures DV. This creates a clear, testable set of data. In correlational designs, only the relation between variables is measured, allowing for various factors to affect the study. However, in correlational designs, the researcher aims to identify the variables and establish a relationship between them.
When using the correlational method, researchers cannot control the extraneous variables and these affect the reliability and validity of the findings. The experimental method allows the experimenter to conduct the study in a controlled environment, preventing the extraneous and confounding variables from affecting the study.
Another difference between the correlational and experimental designs is the sample
size. In experimental design, due to having to conduct the study in a controlled fashion, the sample used generally has to be reduced. In correlational design, the sample size can be large and randomised as various observational methods on different groups of people in different settings can be used.
The experiments usually require various resources, making them an expensive and time-consuming type of research method, while correlations are mainly based on observing and noting the relation between the variables, which does not require extensive amount of funding or time.
13. Give your own example of a between subject’s design.
In between subject’s design, two or more groups in the experiment are put through different experimental conditions. An example of this kind of design would be testing the effect of amount of alcohol consumed on erratic speech. In this experiment one could put participants into different alcohol unit groups and test whether the speech skills vary between groups depending on the amount of alcohol consumed.
14. Consider the following study: A psychologist wants to find out if a mental training course she has devised affects the ability of fire fighters to make decisions in emergency situations. She randomly selects 25 fire stations in the UK. From each of these selected fire stations she randomly selects 10 fire fighters to take part in the study. These 250 fire fighters were then randomly allocated to 2 different groups. One group was sent on a 6 week long mental training course and the other group were not sent on any course. At the end of the 6 weeks both groups were given a test. The test involved each fire fighter being given a simulated emergency scenario and each had to describe what decisions they would make in their scenario to solve the problem. Expert judges (experienced fire fighters) evaluated the participant’s decisions and gave them a score out of 100. The scores for each group were compared and it was found that the group that did the training course scored better than the group that did no course. The difference between the groups was statistically significant.
(a) Identify the population in this study.
The population in the study are the firefighters.
(b) Identify the sampling method that was used.
Random sampling used.
(c) Identify the sample.
The sample in the study are 250 firefighters.
(d) What sort of experimental design is this, i.e., within or between subjects? [1 mark]
Between subject’s design was used.
(e) What is the independent variable (specifically)?
The independent variable is the 6 weeks mental training course given to one of the groups. Second group was the control.
(f) What is the dependent variable (specifically)?
The dependent variable is the score on the tests, measured out of 100.
(g) Describe how you would design this experiment in another way? What are the advantages and disadvantages of this other design?
In this experiment, I would use the within subject’s design. I would administer the test to the group of firefighters before the 6-week mental training course, to establish the baseline, and then after 6 weeks administer the test again to see if there is any improvement. The advantage of this method is that one could see whether the training improves the performance compared to no training, as the group underwent both conditions of the IV. The disadvantage of this method is however the order effect, where the firefighters will know how the test is constructed and the questions within it.
(h) Name one possible threat to the internal validity of this experiment. In other words, can you think of a possible confounding variable that could have affected the results of this study?
The baseline knowledge of firefighters was not established, therefore improved results may be due to knowledge acquired prior to the study.
(I) How could the design be improved so that this confounding variable could be eliminated and the internal validity of the experiment improved?
The researcher could make another test and administer it to both groups, establishing what level of knowledge they have prior to the study.
(j) What does ‘the difference between the groups is statistically significant’ mean, in this specific example?
This means that the difference between the control and experimental group have 5% or less probability to be due to chance.