Research Methods and Design in Psychology

Paul Richardson, Allen Goodwin, Emma Vine

• Publisher: Learning Matters
• Year: 2011
• ISBN:
  9780857254702
• Edition: 1
• Topic: Psychology
• Subtopic: Research & Methodology in Psychology
• Index: Psychology

This accessible introductory text addresses the core knowledge domain of research methods. It provides concise coverage of the central concepts, techniques, problems and debates in this key area, while encouraging a critical approach and developing students? higher level skills.

Activities help readers build the underpinning generic critical thinking and transferable skills they need in order to become independent learners, and to meet the relevant requirements of their programme of study.

The text provides core information on designing psychology research studies with key chapters on both quantitative and qualitative designs. Other chapters look at ethics, common problems, and advances and innovations.

Chapter 1

Designing psychological research studies

Learning outcomes

This chapter will outline the phases and problems associated with the early stage of conducting research. It will take you through the process of finding, defining and designing a research question and highlight the factors that need to be considered.

By the end of this chapter you should:

– understand the stages involved in setting up a research study;

– be able to undertake the process of finding, defining and designing a research study;

– understand the importance of a literature review in the process of conducting research;

– understand the need for a formal operationalisation of the research question;

– be aware of the problems of capturing a measure;

– be aware of the issues of validity, reliability and confounds.

Introduction

The statement that ‘Good researchers are made, not born’ is worth keeping in mind when you approach the design of research studies. Far too often it is assumed that the ability to conduct research is somehow innate or easy. The truth is far more brutal – good research is a process based on acquired knowledge and experience. In order to make this whole process less daunting, we can start by fragmenting the research process into discrete steps. The reasons are twofold. The first is that to approach the process as a whole entity may be beyond our cognitive ability (there are only a limited number of cognitive balls that we can juggle). Secondly, through the categorisation of each stage, we can minimise errors that we may make along the way. It is important to realise that the use of the most sophisticated statistical analysis does not compensate for a badly framed research question or poor methodology, as these invalidate any statistical finding. Always remember that the statistical analysis conducted on the data obtained is only as good as the quality of the data. If the gathered data is flawed, then the findings will also be.

Taking this into account, let us consider the research process as a set of stages outlined below.

• Finding a research question.
• The inevitable literature review.
• Generation of the hypothesis.
• Sampling.
• The operationalisation of the research question.
• Validity, reliability and confounds.

Finding a research question

Without a clearly defined research question, there is no research. The struggle is how to even get to that point. This is where two important aspects of the researcher (you) come in: interest and knowledge.

A good starting point is the reason for the researcher’s areas of interest. It may be that you observed a social situation where something unexpected occurred. What caused that particular outcome? What were the dynamics involved? Alternatively, while reading some recently published research, a question may present itself – something that you read did not make sense. Finally, abandoning personal interest, it could be that you were hired to conduct a piece of research on a predefined question. The sources of inspiration for a research question can be potentially boundless, limited only by the extent to which you wear your ‘researcher’s hat’ when interacting with the world around you. The second aspect – ‘knowledge’ – is the part that requires considerable effort and hard work. The last thing that you want to do is ask a question that has already been answered, and the only way you can be certain of this is to have an up-to-date knowledge of the area. Conducting a literature review is the most effective way of attaining this. In addition, it will also help in writing up the traditional undergraduate laboratory or research report.

The literature review

If the area of interest is new to you, then it may be of use initially to use an undergraduate psychology textbook that deals with the global aspects and themes of the area of interest. This will allow you to get a good feel for the general areas of contention. Importantly within the text, specific references will be made to primary sources of research (the original research written up by the researchers themselves) directly related in a more focused way to your ‘research question to be’. However, there will be a time when you have to enter the world of the online refereed journal.

Academic institutions spend a great deal of money on online peer-reviewed journals, which can be accessed from a variety of online databases (for example, Psychlit, PsychINFO, Science Direct etc.). Interrogation of these online databases is an essential stage yet is also a time-consuming process, which requires identifying the relevant material from that which is irrelevant. However, there are a few tricks that more experienced researchers employ in order to facilitate the gathering of relevant information.

The first is to search for review journal articles. These are journal articles in which experts in a particular field have already reviewed the relevant material, examined the contradictions inherent to any area, and come up with an overall considered, educated and well-argued conclusion. There are two main types of review article – the classic literature review article and the meta-analysis (for a discussion of the advantages of the meta-analytic review, see Rosenthal (1991)). What both have in common is a considerable list of related articles within the reference section that can be plundered to gain more knowledge of the area.

The second trick is the use of keywords. Once you have found a relevant article related to your research interest, you will find, often at the bottom of the abstract, a list of words. These are keywords, which are tags or descriptors related to the article. Related articles will have the same tags or keywords. These can then be used in order to search online databases. The third trick is the use of the citation index. If an article you have identified as relevant to your area of research has been cited by other proceeding articles, then there is a good chance that these later works will also contain material relevant to your area of interest.

Using these three main methods of gathering material, you should be able to gather a considerable amount of relevant material and identify numerous ideas for research. It is also possible that your original research idea may have been answered or that a more interesting question has come to light. Now comes the next step of forming a clearly defined research hypothesis.

The research hypothesis

So what is the research hypothesis? In the simplest terms, a hypothesis is a statement of the predicted outcome of a manipulation. Of course, the manipulation can be an inherent characteristic of the participant (for example, gender, football team supporter or music preference) or an active manipulation (for example, drug dosage, stress manipulation). A hypothesis makes the assumption that there is a specific relationship between the manipulation of the experimental conditions and the observed (or measured) outcome. Formally, there are always two types of hypothesis – the null hypothesis (H_o) and the experimental hypothesis (H_e) (or alternative hypothesis (H_a)). There is always at least one experimental hypothesis for each type of manipulation, otherwise more formally known as the independent variable or factor (see Chapter 4 for further explanation). The null hypothesis, although usually more formally stated, expresses that the manipulation has no effect upon the measured outcome (referred to as the dependent variable). An important thing to note is that we never test an experimental hypothesis but rather reject or fail to reject the null hypothesis. This may seem rather odd until you realise that with the statistical tests that we employ as research methodologists, we are testing the extent to which the outcome of a manipulation (experimental condition) could have happened due to chance. It is only when this is so unlikely as to happen by chance that we refute the null hypothesis and ‘accept’ the experimental hypothesis as the only other credible alternative.

Now, consider the following example. Researchers are interested in courage and (after a great deal of research) were convinced that there was a difference in innate courage between ‘contact’ and ‘non-contact’ sport players. They wanted to test this empirically. In order to state this research idea formally, they generated the following hypotheses:

Null hypothesis: (H_o): There will be no difference in the innate courage between ‘contact’ and ‘non-contact’ sport players.

Experimental hypothesis (H_e): There will be a difference in innate courage between ‘contact’ and ‘non-contact’ sport players.

In the above experimental hypothesis, there is no attempt to state any direction of outcome in relation to how type of sport played reflects innate courage. We refer to this as a two-tailed hypothesis. This is because we do not infer the direction of causality. This is where we could argue that, as true scientists, we keep an open mind to the possibility that the results may be unexpected. There is also a statistical implication of not having a two-tailed hypothesis – in short, it is easier to demonstrate statistical significance if the hypothesis is directional, something that we as researchers need to take into account when reading other research. A directional hypothesis (or unidirectional hypothesis) is one where the direction of the effect is clearly stated. We might therefore write the above hypothesis as:

H_e: Those with a high level of innate courage will play ‘contact’ sports when compared to those who play ‘non-contact’ sports.

Here we can clearly see that we are predicting the exact nature of the outcome of our research. So let us stick to our non-directional, or two-tailed, hypothesis. The next step in our research is to work out how we will measure heroism and courage, and all the pitfalls and fun that might entail. We now need to operationalise our research question with formal operational definitions. But first, let us consider where our participants may come from.

Samples, sampling and populations

With any piece of research, it is impossible to ask everyone in the world to take part to ensure that our results apply to everyone. Therefore, we do the next best thing – we take a sample of individuals from a specified population. The use of ‘population’ as a term may seem odd; what we mean by population is what represents its defining characteristic. For example, the population of psychology students refers to all those individuals (with all their other individual differences) who study psychology. We could also have a sample of a population of left-handed individuals, males, females, psychology students or, even more specifically, female, left-handed psychology students. How and what we define as being the population is dependent on our research question and reflects the characteristics of interest of our pool of individuals from whom we will sample. How we sample, and the number of participants that we recruit, may have implications for the validity and reliability of our research. It is important to make sure that our sampling is unbiased and representative (that all the differences in ability/scores/attitudes due to individual differences are as widely distributed as that of the population from which we have sampled).

The preferred method of sampling is called random sampling. Random sampling is when all individuals within a population have the same chance of being included in the research as a participant. With a sufficient sample size, a random sample has the most likelihood of reflecting the range of data as found in the population as a whole. Although this method is preferable to other methods of sampling, it is often the most expensive and not possible due to financial constraints. A common mistake is to confuse random sampling with random allocation. Random allocation is where the...

Table of contents

• Cover Page
• Title
• Copyright
• Contents
• Acknowledgements
• Series editor’s introduction
• Introduction
• Chapter 1 Designing psychological research studies
• Chapter 2 Ethics
• Chapter 3 Levels of measurement
• Chapter 4 Experimental designs
• Chapter 5 Time course studies
• Chapter 6 Correlational designs
• Chapter 7 Introducing qualitative design
• Chapter 8 Qualitative data collection
• Chapter 9 Qualitative data analysis
• Chapter 10 Mixed methods
• Chapter 11 Common problems
• Chapter 12 Report writing
• Glossary
• References
• Index