Measurement and data analysis

Question 1

What is reliability? What does it contributes?

Answer 1

Reliability is when the results are repeatable when the behaviours are remeasured. It allows meaningful comparison between scores and can understand what the measure means .

Question 2

What is content validity?

Answer 2

Concerns with whetheror not all the contents of the items on a test measures the construct and if it makes sense.

Question 3

what is Criterion validity?

Answer 3

Concerns whether

1. future behaviour can be accurately predicted and is
2. meaningfully related to some other measure of behaviour (produce results similar to other related tests)
3. correlation between test and criterion

Question 4

What is construct validity?

Answer 4

Concerns whether the test measures what is supposed to measure.

Question 5

What is descriptive stats?

Answer 5

It summarises the data from the study or condense a group of scores.
includes measures of central tendency (mean median mode), variability (range, SD and variance) and association

Question 6

What is inferential stats?

Answer 6

Draw conclusions from data (in exp) to population

Question 7

Know what is relative risk.

Answer 7

A measure of the risk of a certain event happening in one group (exposed group) compared to the risk of the same event happening in another group (unexposed group)

Question 8

What is Type 1 error and type 2 errors?

Answer 8

1. Reject Null hypo (but wrong)

2. Don’t reject Null hypo (but wrong)

Question 9

What is systematic variance?

Answer 9

An identifiable factor (e.g. variable of interest or some other factor) that you failed to control adequately.

Question 10

What is error variance?

Answer 10

Nonsystematic variability due to individual differences o r random error (temp.. etc)

Question 11

What is effect size?

Answer 11

Provides an estimate of the magnitude of the difference of sets of scores while taking into account the amount of variability of the scores.

Question 12

What is an advantage of effect size?

Answer 12

Researchers are able to arrive at a common metric for evaluating a diverse set of experiments.

Question 13

NON Overlapping confidence interval indicate a substantial difference between the two conditions of study.

Answer 13

NIL

Question 14

What is power of a test affected by?

Answer 14

1. Size of treatment (effect size)
2. alpha level
3. size of sample

Question 15

What error increases with power decreases?

Answer 15

Type 2 error

Question 16

What are the ways to assess reliability?

Answer 16

1. Test and retest - correlation of scores between 2 occasions
2. Parallel form - Correlation between equivalent forms of the test that have different items.
3. Internal consistency - correlation of items within test (alpha)

Question 17

What is sensitivity and specificity?

Answer 17

1. Sensitivity - refers to the number of correctly identified positives.
2. Specificity - refers to the number of correctly identified negatives.

Question 18

What is ambulatory assessment?

Answer 18

Computer assisted methodology for self report, behaviour records etc (Handys, handheld etc. Subjects use devises to ans questionnaires etc)

Question 19

Low reliability will NOT have high validity.

Possible to have high reliability and low validity, low reliability and low validity or high on both.

Answer 19

NIl

Question 20

What are parametric tests?

Answer 20

ANOVA, T test. They assume sample’s parameters and distribution same as the population and we use it when data are normally distributed.

Question 21

What are non parametric tests?

Answer 21

chi-square, Fisher’s exact test and the Mann-Whitney test.We don’t assume anything about the population’s parameters and distribution and we use it when the data are non-normally distributed.

Question 22

What are the assumptions underlying ANOVA one factor design?

Answer 22

1. Individual differences and errors of measurement are independent from group to group (by randomly distributing subjects to group)
2. Individual differences and errors of measurement are normally distributed within each group
3. The size of variance in the distribution of individual differences and random errors is identical within each cell

Question 23

How do the scales of measurement affect the selection of the test statistic? (state 2 examples)

Answer 23

1. Nominal & Ordinal -> Non parametric

2. Interval and ratio -> parametric

Question 24

Difference between parameter and statistic?

Answer 24

Parameter: any statistical characteristic of a population (e.g. population mean, median etc)
Statistic: any statistical characteristic of a sample (e.g. sample mean, sample median etc)

Question 25

What is the sum of differences also known as?

Answer 25

Deviation scores

Question 26

Why are the disadvantages of using variance?

Answer 26

Highly susceptible to outliers and data needs to be symmetrical. Usually used with advanced stats like Multiple regression and ANOVA

Question 27

What are some of the problems of test and re-test?

Answer 27

1. Regression towards mean
2. Hard to replicate exact conditions
3. Subjects may become better as they are used to the task

Question 28

How do u measure correlation and what is the problem with this?

Answer 28

1. Measure r

2. Sensitive to outliers