Methods (general)

Question 1

Why is significance not equivalent to the strength of the correlation?

Answer 1

Because significance is affected by sample size

Question 2

What are some strengths of case studies?

Answer 2

They offer detailed, rich insight into real life cases. They can generate hypotheses for future study and challenge current theories.

Question 3

What are some weaknesses of case studies?

Answer 3

Inappropriate to generalise from one-off cases. Interpretation of such cases can be subjective. They rely on truth and accuracy from the individual.

Question 4

What are some strengths of content analysis in general?

Answer 4

It avoids ethical issues involved in obtaining permission as much of the communications are publicly available. It has high external validity. It is flexible in that it can produce qualitative or quantitative data.

Question 5

What are some weaknesses of content analysis in general?

Answer 5

The communications are studied outside of the context they occurred so may be misinterpreted. It may suffer from a lack of objectivity.

Question 6

Evaluate the use of quantitative content analysis

Answer 6

Strengths: objective, produces clear statistical patterns
Weaknesses: Inflexible, lacking in detail

Question 7

Evaluate the use of qualitative content analysis

Answer 7

Strengths: detailed, preserves depth of data
Weaknesses: subjective, time consuming

Question 8

Explain the test-retest method of assessing reliability

Answer 8

The same test is administered to the same participant on different occasions. An appropriate amount of time should be allowed between the test and retest - not so much that the participant themselves has changed, but not so little that they recall their answers to the test. If the test is reliable it should produce the same results both times (this is tested by correlating the two sets of results).

Question 9

How can the reliability of a questionnaire be improved?

Answer 9

If the test-retest method shows poor reliability, complex or ambiguous questions should be removed or rewritten.

Question 10

How can the reliability of an interview be improved?

Answer 10

The same interviewer should be used for all interviews, and if this is not possible interviewers should all be properly trained in the same way. Structured interviews are most reliable.

Question 11

How can the reliability of experiments be improved?

Answer 11

Through standardisation and making the experiment replicable

Question 12

How can the validity of experimental research be improved?

Answer 12

Using a control group, standardisation, and single/double blind procedures

Question 13

How can the validity of questionnaires be improved?

Answer 13

By using a lie scale, and insuring anonymity

Question 14

How can the validity of observations be improved?

Answer 14

By doing a covert observation, and ensuring that behavioural categories are not too broad, overlapping, or ambiguous

Question 15

How can the validity of qualitative methods (e.g. case studies, interviews) be improved?

Answer 15

Demonstrating that the researcher’s interpretations match the participant’s (interpretative validity), using a number of different sources as evidence (triangulation)

Question 16

What are the 3 criteria for using a parametric test?

Answer 16

1. Interval level data, 2. normal distribution, 3. homogeneity of variance (similar spreads in each condition)

Question 17

Order of the statistical tests in the table (left to right)

Answer 17

Chi, sign, chi, mann, wilcoxon, spear, unrelated, related, pear

Question 18

Order of levels of measure in the statistical test table (downwards)

Answer 18

Nominal, ordinal, interval

Question 19

Order of type of investigation in the statistical test table (left to right)

Answer 19

Unrelated design, related design, correlation

Question 20

Which measures of central tendency and spread are related to nominal level data?

Answer 20

Mode and n/a

Question 21

Which measures of central tendency and spread are related to ordinal level data?

Answer 21

Median and range

Question 22

Which measures of central tendency and spread are related to interval level data?

Answer 22

Mean and standard deviation

Question 23

How do you remember what is meant by a type I error?

Answer 23

I -> L -> Lenient i.e. it is too easy to get a significant result (significance level too high)

Question 24

How do you remember what is meant by a type II error?

Answer 24

II -> 2 -> S -> Stringent i.e. it is too difficult to get a significant result (significance level too low)

Question 25

Explain how to rank data in statistical tests involving ordinal data

Answer 25

1. Write down all the different values of data which occur in order
2. Next to each value, write the number of times it occurs in the sample
3. Next to the frequencies, write the ranks which these data will take (e.g. frequency 3 could be ranks 4,5,6)
4. Find the means of the groups of ranks
5. Give each piece of data its corresponding mean rank

Question 26

What is the order of the sections of a scientific report?

Answer 26

Abstract, introduction, method, results, discussion, referencing

Question 27

State the 5 steps of theory construction and hypothesis testing

Answer 27

1. observations made, 2. tentative theory proposed, 3. specific hypothesis deduced, 4. hypothesis tested, 5. theoretical development

Question 28

Briefly outline the claim that Psychology is not a science

Answer 28

Kuhn claimed this because Psychology has too much internal disagreement and lacks a paradigm, and some areas of Psychology are unfalsifiable, subjective, and not replicable.

Question 29

Briefly outline the claim that Psychology is a science

Answer 29

Psychology adopts the empirical method, theory construction, and hypothesis testing, and some (most) areas are falsifiable, objective, and replicable.

Question 30

Explain the inter-observer method of assessing reliability

Answer 30

The two observers do a pilot study to check that they are independently applying the behavioural categories in the same way. If their observations differ they discuss and modify the categories. Then they observe the real study, and their reliability is tested by correlating their observations.

Question 31

How can the reliability of observations be improved?

Answer 31

By making sure the behavioural categories are operationalised, not overlapping, and cover all possible behaviours. Also by testing inter-observer reliability before conducting the observation.