Correlational Studies

9 Key excerpts on "Correlational Studies"

• 

  eBook - ePub

  Laboratory Psychology

  A Beginner's Guide

  • Julia Nunn(Author)
  • 2018(Publication Date)
  • Psychology Press

    (Publisher)

  allocate a group of infants to a condition where they were deprived of their mothers. Of course an experimental design permits greater control over the influence of other variables but the advantages of active manipulation of variables carry with them a moral duty to act responsibly.

  SUMMARY

  Correlational Studies allow analysis of behaviour that has been passively observed—rather than collected under experimentally rigorously controlled conditions.

  • Determining the cause of a correlation between variables requires special care. If two variables A and B are correlated then there are three possible causal interpretations:
    • (i) A may be the cause of B;
    • (ii) B may be the cause of A;
    • (iii) Some other variable may cause A and B.
  • Correlational Studies can be used to investigate behaviour occurring outside the laboratory which it might not be ethical or practically possible to observe under laboratory conditions.

  Experimental and correlational approaches: A comparison

  Whereas the experimenter is interested in the variation he creates by manipulating variables, the correlator is interested in the variation already existing between individuals and/or groups of individuals within or across cultures. The correlator’s aim is to understand the variance produced by nature. For the experimental psychologist individual differences in the subjects of the study are often seen as a problem rather than a focus of interest. A large amount of variation obtained between subjects within a treatment condition is seen as an indication that the experimenter has not controlled the experiment tightly enough. In experimental studies individual variation is termed “error variance” and is one source of variation that the experimenter will deliberately try to eliminate by matching subjects or using homogeneous subject groups.

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Research Methods and Statistics in Psychology

  • S Alexander Haslam, Craig McGarty(Authors)
  • 2018(Publication Date)
  • SAGE Publications Ltd

    (Publisher)

  9 Examining Relationships between Variables: Correlation

  Key goals for this chapter

  1. Introduce correlation – a procedure for examining relationships between variables.
  2. Introduce Pearson’s r – a statistic that quantifies the nature and strength of the linear relationship between two variables.
  3. Explain how the results of correlational analysis should be interpreted, with particular reference to the difference between correlation and causation.

  A lot of what we have talked about in the previous two chapters relates to comparisons between means, and comparing means is what psychologists typically do when they use experimental methodology. However, in a great deal of research investigators confront another interesting question: What is the relationship between two variables? For example, how is stress related to heart disease? How is socio-economic status related to mental health? How is personality related to the judgements people make?

  This type of question can be addressed by experiments (as in the previous chapter, where we used an example of the relationship between the amount of physical contact and attraction), but is more typically examined in surveys. In surveys the researchers collect information about variables where there may be many different values for each variable (in the attraction study, for example, lots of different levels of contact, not just two, and lots of different levels of attraction). Surveys also often measure two or more variables, each with multiple levels. We can contrast this with the case of a t-test where we have an independent variable with just two levels (a categorical variable) and a dependent variable with multiple levels.

  In survey research we often want to know whether two variables vary together. In other words, we want to determine whether there is an association between the variables. This is quite different from the example of the within-subjects t-test, where we are interested in the question of whether there are informative differences between participants’ scores on different measures. In this chapter we want to look at pairs of scores to see whether high scores are consistently associated with high scores (and low scores with low scores) or whether high scores are consistently associated with low scores. Where there is a relationship between variables that takes either of these two forms we say there is a correlation

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Research Methods and Design in Psychology

  • Paul Richardson, Allen Goodwin, Emma Vine(Authors)
  • 2011(Publication Date)
  • Learning Matters

    (Publisher)
• We have specifically pointed out that this correlational design is used exclusively to assess the degree of association between two (or more) variables. It is not looking at differences between scores, and therefore produces no information about the cause-and-effect relationship between these variables. However, it is true to say that without a significant correlation coefficient, then no relationship exists between the variables. What are the implications of this statement?
• Any conclusions that you may draw from a study involving the association between two variables might be at risk due to the influence of other variables. Therefore, researchers will often measure other variables and partial their effects out, i.e. attempt to remove their influence on the other variables. If you conducted such a study, how would you decide which other variables to collect – what would you need to consider in order to select such variables?
• An extremely useful extension of researching the degree of association between variables is the ability to predict future scores. This may be a new concept to you in research methodology, but what do you think are the possible benefits of this approach? Can you foresee any limitations?

Summary: what you have learned

In this chapter, you were introduced to another type of design often used in research – the correlation. However, unlike the experimental designs presented in the preceding chapters, studies that use correlation are not investigating the causal relationship between variables – no variables are being actively manipulated. Instead, this research design helps us to investigate the degree of association between two (or more) variables. People often confuse correlational designs with repeated-measures (within-participant) designs. This is because in both designs participants are tested on the tasks on numerous occasions; however, correlated designs do not require the same tasks to be completed – it is possible to study the association between different tasks or variables!
We discussed how the associations between two variables can be visually represented by using a scatterplot. This helps to demonstrate any linear trends in the data. This association can be measured with a correlation coefficient – a figure that lies somewhere between ±1 and 0. This serves two purposes: it shows the direction of the relationship (positive, negative or none) and also the strength of this relationship (weak, medium or strong).