Scales of measurement, reliability, validity and central tendency WK2

Question 1

Differ between discrete and continuous data

Answer 1

Discrete
-No intermediate values between data values
-Nominal and ordinal measures=discrete
-Cannot be infinite decimal between points of scale

Continuous
-Intermediate values between data values
-Common in interval and ratio scales
-Can be infinite decimals between points of scale

Question 2

Provide summary of the types of reliability

Answer 2

Internal consistency
Test-retest reliability
Inter-rater reliability

Question 3

Describe internal consistency

Answer 3

Consistency among items in a measure
-Test takers responses to all items on a scale/subscale should be similar/consistent

Question 4

Describe test-retest reliability

Answer 4

Consistency overtime
-Test takers responses to items at time 1 should be similar to responses at time 2

Question 5

Describe inter-rater reliability

Answer 5

Consistency across observers/raters
-Multiple observers/raters should provide similar accounts of same event/occurrence/behaviour

Question 6

Differ between reliability and validity

Answer 6

reliability: consistency of measure

validity: Extent to which concept or measurement is well founded and corresponds accurately with world

Question 7

Provide summary of the types of validity

Answer 7

Face Validity
Content Validity
Criterion Validity
Construct Validity

Question 8

Describe face validity providing example relating to extraversion/sociability measure

Answer 8

Looks like it is measuring X
e.g. Items appear (to naïve observer) to be measuring extraversion/sociability

Question 9

Describe content validity providing example relating to extraversion/sociability measure

Answer 9

Samples the full breadth of X
-Items cover all aspects of extraversion e.g. Talkative, adventurous

Question 10

Describe criterion validity providing example relating to extraversion/sociability measure

Answer 10

Tests correspondence with concrete outcome; measured now (concurrent) + in future (predictive)
e.g. Scores on measure predict current base rate cortical arousal levels; as well as future social behaviour ate parties/work

Question 11

Describe construct validity providing example relating to extraversion/sociability measure

Answer 11

Related to other measures of X (convergent), not related to measures of Y and Z (discriminant)
e.g. People with high score on measure also score highly on NEO-PI extraversion subscale but not related to Beck Depression Inventory

Question 12

What is meant by scales of measurement

Answer 12

More properties possessed by data, more flexible our options are (more manipulations performed to evaluate data)

Question 13

Provide summary of the types of scales (try with table)

Answer 13

Nominal/categorical
Ordinal
Interval
Ratio

Question 14

Describe nominal/categorical scales of measure providing examples

Answer 14

-Most limited scale in quantitative research
-No property of magnitude
-Can’t perform operations
-Data with only 2xcategorical options=dichotomous data (e.g. Smoker vs non-smoker)

e.g. Biological sex, Political preference, Race/ethnicity, Eye colour

Question 15

Describe ordinal scales of measure providing examples

Answer 15

-Property of order but values do not represent magnitude
-Cannot perform operations (e.g. subtraction addition)
-May represent categories

e.g. Placing in competition, Low drink medium drinker high drinker, Nando’s sauce peri-ometer

Question 16

Describe interval scales of measure providing examples

Answer 16

-Property of magnitude but not relative to magnitude
-Meaningful (equal) differences between intervals
-No true zero point (can have negative values)
-Zero does not represent absence of property
-Can perform some operations (subtraction, addition)

e.g. Temperature, 12-hour clock

Question 17

Describe ratio scales of measure providing examples

Answer 17

-Property of magnitude (including relative)
-Meaningful (equal) interval differences
-True zero point
-Zero represents absence of property
-Can perform many operations (sub, addit., multip., division)
-Most useful in analysis

e.g. Reaction time, Weight, Exam score, Height

Question 18

What is the relationship between reliability and validity

Answer 18

A measure can be reliable but not valid; however, an unreliable measure cannot be valid

Question 19

Describe what is meant by measures of central tendency

Answer 19

One number summaries (more/less the simplest form of stats)
-Mean: sum of values in data set divided by number of values
-Median: midpoint of dataset
-Mode: most frequent value in dataset (not popular in psychology)

Question 20

How to determine use of mean or mode

Answer 20

-Depends on shape of data
-Mean useful when symmetrical distributions
-Problematic using mean when data not symmetrical
-Better to report median when data is skewed

Question 21

Summarise the measures of variability with definitions for each

Answer 21

-Range: difference between smallest+largest value in dataset
-Interquartile range: range between middle 50% of data
-Variance: sum of squared deviations of each data point from mean
-Standard deviation: square root of variance
-Outlier: data point differs significantly from other observations

Question 22

Differ between hypothesis and theory

Answer 22

theory is broader/bigger testable statement.

Hypothesis is tentative statement