L2: individual differences and correlations

Question 1

what are: Measures of dispersion and correlation

Answer 1

Range
Standard deviation
Variance

correlation:
- standardised representation of the relationship between two variables
- from -1.0 to 1.0

Question 2

composite variables
and variance

Answer 2

• test scores tend to be based on the sum of two or more items
• sum score are known as composite scores
• The variance of a composite score is a function of:
  1: the variance associated with the individual items, and
  2: the correlations amongst the items

Question 3

Binary items

Answer 3

• most frequently encountered in achievement type test eg exams
• responses are score as 0 or 1, correct or incorrect

Question 4

2 most common interpretations of test scores
relative and abstract

Answer 4

1. relative based on analysis of data to do this we need to:
   - make reference to an entire distribution of scores (know the mean)
   - identify where the individual falls within that distribution (know the SD)
2. Abstract: based on theoretically relevant characteristics of the body of research that support test scores as valid indicators of a psychological construct

Question 5

Z scores
T scores
Percentile ranks

Answer 5

z score
- have a mean of 0 and an SD or 1
- therefore some can fall in the negative range

T score
- z scores rescale so there’s no negative range
- mean is always 50 and SD is always 10

ranks
Percentile ranks indicate the percentage of scores that are equal to or below a specific test score.

Question 6

reference samples, what does it help with?
representativeness
probability sampling & random sampling
non-probability sampling

best sample size

Answer 6

• It’s also known as a ‘normative sample’
• helps us compare scores
• The quality of normed data depends on the representativeness of the sample.
• In IQ testing, you would want a reference sample that consists of people from a wide range of walks of life, for example, gender, education, socioeconomic status, etc.
• Good IQ tests make sure that their reference sample corresponds to that country’s Bureau of Statistics information
• Probability samples use procedures that ensure a representative sample eg 20%, 20%, 20% or X X X type of people
• Random sampling is a type of probability sample, as you would expect a random sample from the population to be representative of the population. In reality, is not achievable as there is never all equal chances of being picked
• non-probability sampling: In most testing, there is self-selection
• The best sample size is 500

Question 7

Closed-ended questions
advantages and disadvantages

Answer 7

asks respondents to choose from a fixed set of response alternatives
eg: yes/no questions, multiple choice questions, true-false, numerical rating scale (Likert)

Advantages
- Fast and easy for respondents
- Restrict the range of responses
- Helps keep people on the topic of the question
- Closed-ended questions can also remind people of things they otherwise would have forgotten.

disadvantages
- Participants are restricted to responses that may not accurately reflect their feelings or attitudes about the topic.
- Closed-ended questions may force people to express an opinion when they don’t actually have one on the topic
- Assuming none of the alternatives is ‘Not applicable’ or ‘No opinion’.
- In some contexts (job satisfaction surveys), closed-ended questions may cause frustration in the respondents, if they are made to feel that they were no allowed to express their true feelings.

Question 8

open-ended questions
advantages and disadvantages
alternatives

Answer 8

open-ended questions do not provide respondents with nay response alternatives. they respond in their own words
eg long or short answer questions

advantages
- very informative/ lots of depth
- For sensitive topics (e.g., drug use), it can be more useful for people to respond in a less constrained way than in a fixed-answer format.
- Allow people to express their precise feelings/attitudes, without forcing them to endorse an inaccurate closed-ended question fixed answer.
- Gives people the sense that their opinion matters.

disadvantages
- They take a relatively long time for the respondent to complete and analysis is costly for time too
- People may get tired/bored by the end of the survey and answer the last questions perfunctorily.
- Some people will be so uninterested in the survey that they will give very poor open-ended responses.
- give poor quantitative data
- All too often, people “unload” rather than provide responses to the question.

alternative
- An alternative to truly open-ended questions would be to ask people to list the top 3 reasons for something.
- For example, you could ask people to list the top 3 reasons why people quit the company

Question 9

point rating scales
advantages and disadvantages

Answer 9

• The most widely used response format
• most widely used is the Likert scale.
• Respondents provide ratings indicating how strongly they feel positively or negatively about an issue.
• Likert scales use words to ‘anchor’ the numerical ratings.
• Can be as few as 2 and as many as 11 points.
• Typically, respondents cannot make the distinctions between adjacent points in a rating scale with 11 or more points (too many points for response reliability)
• Most surveys use between 4(too little) and 7(the best) points.

advantages
- Likert scales are efficient to create (i.e., basic ones)
- Most respondents have experience with Likert scales.
- Using the same response alternatives for numerous items allows for the collection of a large amount of information in a relatively short period of time.
- Likert scales yield numerical scores that can be analysed statistically.

disadvantages: midpoints
- It is debatable whether a test developer should use ‘Neither Agree/Nor Disagree’ in a response scale.
- It may be better to force people to make a decision one way or the other
- Some people have personality types that lend themselves to responding ‘Neither Agree/Nor Disagree’.
- They are not the type to express an opinion (even though they may have one).
- Sometimes, it is unavoidable to use a response such as ‘not applicable’, but this is not a midpoint

verbal anchors
- hard to determine what exactly almost always and seldom means for the participant

Question 10

A good test will have:
negatively keyed items
double-barrelled items
simple lanaguage

Answer 10

• include a mix of positively keyed and negatively keyed items.
• By including negatively keyed items, you prevent people from getting into a response set (e.g., acquiescence).
• The endorsement of a positively keyed item entails that the person is affirming the attribute of interest.
• By contrast, the endorsement of a negatively keyed item entails that the person is denying or negating the attribute of interest
• A double-barrelled item asks two or more questions in the same item that may be expected to be answered differently by at least some respondents.
• Anytime you see the word ‘and’ or ‘or’ in an item, you have likely encountered a double-barrelled item (that’s bad).
• Keep the items simple and focused on one idea.
• use language that requires an eighth-grade level or lower.
• Avoid slang terms, colloquiums and jargon
• This is true even if it is a questionnaire for adults.
• item should be short: more than 1 comma means too long