RESEARCH METHODS IN PSYCHOLOGY ASSIGNMENT-

This assignment consists of three parts.

PART 1: RELIABILITY

Two researchers, Matt and Ian, were interested in studying cooperative behaviour between two children playing with toys in a large room at a kindergarten. They first agreed on a definition of cooperative behaviour: 'How much a child cooperates with another child'. They then used a partial interval time sampling method. In this methodology, they observed, independently of each other but at the same time, the children playing for 5 seconds and then, in the next 5 seconds, they recorded what they observed. They followed this procedure of 5 seconds observing/5 seconds recording over approximately a 2-minute period. This period is 'Observation 1'.

They then repeated this procedure for another 2-minute period (Observation 2).

Below is a record of what Ian and Matt observed at Observation 1, and what Ian also observed at Observation 2. You will see that there were 10 intervals of time in which they recorded behaviour. X indicates that co-operative behaviour was observed in that interval.

1. To what extent did Matt and Ian agree that cooperative behaviour occurred during Observation 1? And, what sort of reliability is being assessed here?

To answer this question, calculate and write down the point-by-point agreement ratio using the following formula:

Inter-observer agreement = no. agreement on occurrences of behaviour x 100 /no. agreement on occurrence + no. disagreements on occurrence

[Agreement refers to when an X appears in a corresponding interval. For example, in interval 1 at Observation 1, there is an X for Ian, but none for Matt. So, that is a disagreement. At interval 2, however, both Ian and Matt have an X, so that is an agreement. When two corresponding boxes don't have an X, that's not an agreement or a disagreement, so you ignore it].

2. How well did Ian and Matt measure what they thought they were measuring? Briefly explain your answer.

3. How consistent was Ian across the two Observation periods? Use the same formula as above and write down your result. In addition, tell us what sort of reliability is being assessed here.

4. How do the percentage agreements that you calculated for questions 1 and 3, above, compare? What would you conclude? How would you explain these results?

5. If Ian and Matt were going to do the same task all over again what would you advise them to do, if anything, to improve the reliability of their measure?

PART 2: VALIDITY AND RELATED CONCEPTS

Recently, a team of research psychologists decided that the world needed a new measure of happiness. So, they developed two new measures and conducted a study to test the validity of the new measures. Their aim was to find out which of their new measures they should use in subsequent research.

Each of the new measures consisted of 5 items. Each employed Likert-type scales for responding, where 1 = strongly disagree; 7= strongly agree.

The researchers administered the new scales, plus several other measures, to 320 stay at home parents in Adelaide. The study was conducted online. The mean age of the sample was 27 (SD = 5.1). There were 240 women and 80 men. Of these, the vast majority of participants identified as White Australians (94%).

Below are some of the results of the study.

First, the Cronbach's alpha for the internal reliability of each of the new measures was .85 for New Happiness Measure 1, and .52 for New Happiness Measure 2.

Second, here are correlations between the new happiness measures and some other measures.

*** p< .05

Here is some basic information about the additional measures that were employed:

• All items in the measures mentioned below are measured on 1-7 scales where 1 = strongly disagree; 7 = strongly agree, unless otherwise indicated.
• The Oxford Happiness Scale (Hills & Argyle, 2001) is a widely-used existing measure of happiness. It consists of 8 items (e.g., 'I feel that life is very rewarding').
• Optimism was measured with a single item, 'I am optimistic about the future'.
• Satisfaction with Life Scale (Diener et al., 1985) is a widely-used measure of wellbeing. It consists of 5 items (e.g., 'In most ways my life is close to ideal').
• Self-esteem was measured with Rosenberg's (1977) self-esteem scale. There are 10 items in the scale, e.g., 'I feel that I have a number of good qualities'
• Negative affect was measured with five items taken from the PANAS (Positive and Negative Affect Scales; Watson et al., 1988) (e.g., 'Thinking about yourself and how you normally feel, to what extent do you generally feel upset?')
• Age was measured by asking participants how old they were.
• Choice was measured with the combination of two items, 'I feel like I have control over the decisions I make in my life' and 'I am always able to choose what I do in my life'.
• Relationship quality was measured with the 15-item Relationship Investment Scale (Rusbult et al., 1998). An example item is 'I feel satisfied in my relationship'.
• Prosocial behaviour: At the end of the online survey, participants were provided with a link to the website of the charity, Oxfam. Participants who clicked on the link were coded as being prosocial; participants who did not click on the link where coded as being 'non pro-social'.

Third, the researchers tested for differences on marital status. Below is a chart indicating the distribution of married vs non-married participants on happiness as measured by NHM1.

On the basis of the information provided, please answer the following questions.

1. What research design was employed in this study? Briefly justify your answer.

2. State three constructs that were measured. For each of these three constructs, state how each of these constructs has been operationalized.

3. What type of data is yielded by the measures of happiness?

4. What hypothesis would you make for the relation between relationship quality and happiness? To help you answer this question, a reading has been provided in the Major Assignment folder (see Myers, 2000). Based on the reading, in around 150-200 words, explain why you would expect this hypothesis to be supported.

5. What analysis was conducted to test this particular hypothesis?

6. Look at the correlations between each of the happiness measures and age. What do you think might explain these non-significant results?

7. Who seems to be happier, people who are married or not married? Briefly justify your answer, based on the results.

8. For the analysis in Question 7 [above], what is the predictor/IV and what is the outcome/DV? Also, what sort of data is yielded by the predictor/IV, and what sort of data is yielded by the outcome/DV?

9. The researchers found a significant correlation of .55 between optimism and happiness.  Can we conclude that being optimistic causes people to be happier? Why or why not? Justify your answer.

10. Think about the correlations found between choice and the happiness measures. One conclusion people might draw is that having choice increases happiness. However, there could also be a third variable explanation for this relation. Identify a logical possible third variable (it may not have been measured in this study). Be sure to explain why your proposed third variable could provide an alternative explanation for the relation.

11. The researchers examined relations between each of the new measures of happiness and prosocial behaviour. They had expected that happier people would be more likely to act prosocially. What type of validity is being assessed here? Briefly explain why you say this.

12. There is something problematic with the way that 'prosocial behaviour' was operationalized. What do you think it is?

13. To what extent could you generalize the results of this particular study to a wider population? Be sure to justify your reasoning.

14. Which of the new measures is most reliable? Briefly explain your answer.

15. Which of the new measures possesses stronger construct validity? Briefly explain why you think this.

16. Which new measure of happiness would you recommend that researchers use? Be sure to justify your decision.

PART 3: DESIGN AN EXPERIMENT

You will present a proposal for an experimental design to solve the following research problem. Note that you are not required to do a literature search to justify your design, you only need to use the information provided within this document. Finally, you are not expected to actually do this study; it's a proposal only.

Your job is to test the hypothesis that people who talk on their mobile phones while driving are more likely to have an accident.

Your proposed study should be conducted in a laboratory environment. To help you imagine that space, there is a waiting room and a large room with a driving simulator. You also have access to research assistants and money to develop computer programs and pay for any electronic equipment you think you might need. Your participants will be first year students participating for course credit. Your proposed study should be ethical (basically, no-one should get hurt or upset) and sensible.

Now, answer the following questions as they relate to your research proposal:

1. What are your independent and dependent variables?

2. What are the two levels (i.e., conditions) on your independent variable?

3. Describe how you will do the study, making sure you demonstrate your understanding of what is required to conduct a true experiment. In particular, explain how you will manipulate your IV, and what your participants will do (or experience) in the two conditions. Be specific and be detailed! In addition, if you have done something in particular to control for potential confounds, it would be in your interest to point out what you have done.

4. What individual differences might be relevant in this study? How will you ensure that such individual differences do not have an unwanted influence on your results?

5. How you will measure the dependent variable?

6. What statistical analysis would you conduct to test your hypothesis? Why do you say this?

1500 words

8 references