1/203
Looks like no tags are added yet.
Behavioural measurement
The measurement of behaviour either because the behaviour itself is of interest or because the behaviour is being used to assess an unobservable psychological attribute.
What are the 2 purposes of behavioural measurement?
Measure a specific behaviour because it is interesting in its own right. Observe behaviour to assess an unobservable psychological attribute.
Hypothetical construct
A theoretical concept used by psychologists to explain differences in people's behaviour.
Latent variable
Another term for a hypothetical construct, meaning an underlying psychological characteristic that cannot be directly observed.
Operational definition
The procedure used to measure a hypothetical construct.
Psychological test, according to Lee Cronbach
A systematic procedure for comparing the behaviour of two or more people.
What are the 3 components of Cronbach's definition of a psychological test?
Tests involve behavioural samples. Behavioural samples are collected systematically. The purpose is to detect differences between people.
Interindividual differences
Differences in behaviour or psychological characteristics between different people.
Open-ended test
A test in which people can provide whatever response they want to a question.
Closed-ended test
A test in which people choose among alternative answers that are provided.
Criterion-referenced test
A test used to make a decision about a person's skill level by comparing performance with an established criterion or standard.
What 2 groups does a criterion-referenced test create?
People whose performance exceeds the criterion. People whose performance does not exceed the criterion.
Norm-referenced test
A test used to determine how one person's performance compares with the performance of others.
Reference/normative sample
A sample of people who complete a test and are thought to be representative of a broader population.
Criterion-referenced vs. norm-referenced
Criterion-referenced = compare a person with a predetermined standard. Norm-referenced = compare a person with other people.
Speeded test
A time-limited test in which people are generally not expected to complete the entire test.
How are speeded tests scored?
By counting the number of questions answered during the allotted time.
Power test
A test that is not time-limited and contains test items that range in difficulty.
How are power tests scored?
By counting the number of correct answers.
Reflective/effect indicator
An observable indicator that is viewed as being caused by an underlying latent construct.
Formative/causal indicator
An observable variable that contributes to an underlying latent variable rather than being caused by the latent variable.
Reflective vs. formative indicators
Reflective = construct -> indicator. Formative = indicators -> construct.
Psychometrics
The science concerned with evaluating the attributes of psychological tests.
What are the 2 foundations of psychometrics?
Psychological testing and measurement. Development of statistical concepts and procedures.
Differential psychology
The study of individual differences.
What areas are commonly associated with differential psychology?
Intelligence, aptitude, and personality.
Participant reactivity
When knowing that one is being measured influences the psychological process or behaviour being measured.
Demand characteristics
Changes in participants' behaviour or responses based on what they think the researcher wants to find.
Social desirability
Changing behaviour or responses in an attempt to create a favourable impression.
Observer/scorer bias
When a researcher's own expectations or biases affect measurement or scoring.
Composite score
A single value that combines multiple test scores, subtests, or questionnaire items to measure a broader psychological construct.
Score sensitivity
A measure's ability to discriminate between meaningful amounts of the dimension being measured.
Scaling
The way numerical values are assigned to psychological attributes.
Property of identity
The property requiring that people within a particular category are identical with respect to the feature represented by that category.
What are the 2 requirements for categories under the property of identity?
Categories must be mutually exclusive. Categories must be exhaustive.
Mutually exclusive categories
Categories that do not overlap, meaning an observation cannot simultaneously belong to conflicting categories.
Exhaustive categories
Categories that provide a classification option for every person or observation being classified.
Under the property of identity, what do numerals represent?
Sameness versus difference or category membership. They do not have quantitative significance.
Property of order
The ability to arrange values in ascending or descending order based on the amount of an attribute.
What does the property of order tell us?
Relative rank, but not the degree or magnitude of differences between ranks.
Property of quantity
A numerical property that provides information about the magnitude of differences between people.
Other terms for quantitative measurements involving real numbers
Scalar, metric, or cardinal measurement.
Absolute zero
A zero point indicating that the measured attribute has no existence.
Arbitrary zero
A starting point selected through convention or convenience that does not indicate the total absence of the attribute.
What does a z-score of 0 mean?
The score is equal to the average of the set of scores.
Absolute zero vs. arbitrary zero
Absolute zero = complete absence of the attribute. Arbitrary zero = conventionally selected starting point.
What are the 3 ways a unit of measurement can be arbitrary?
Unit size is selected by humans. Some units are not tied to one particular type of object. Some physical units can be used to measure different features/objects.
Additivity
The requirement that the size of a measurement unit remains constant across measurement conditions.
Conjoint measurement
The quantitative assessment of qualitative attributes based on numerical values assigned to different contributing attributes.
What is the basic idea behind conjoint measurement?
Break an overall judgment or decision into the different factors contributing to it.
When does counting qualify as measurement?
When counting is used to represent the amount of a particular feature or attribute.
What are the 4 scales of measurement?
Nominal, Ordinal, Interval, Ratio.
Nominal scale
A scale measuring qualitative data by sorting observations into unordered, labelled categories.
What numerical property does a nominal scale have?
Identity.
Ordinal scale
A scale that ranks people according to the amount of an attribute they possess without determining exactly how much of that attribute they have.
What numerical property does an ordinal scale have?
Order.
Interval scale
A scale in which data are ordered, distances between values are equal and meaningful, and the zero point is arbitrary.
What are the 3 main properties of an interval scale?
Ordered values. Equal, meaningful intervals. Arbitrary zero.
Can interval scales be interpreted multiplicatively?
No. For example, 80°C cannot be interpreted as twice as hot as 40°C because zero is arbitrary.
Ratio scale
A measurement scale with ordered values, equal intervals, and an absolute/true zero.
What are the 3 main properties of a ratio scale?
Ordered values. Equal intervals. Absolute/true zero.
What mathematical interpretations do ratio scales allow?
Both additive and multiplicative interpretations.
Measurement scales from lowest to highest
Nominal -> Ordinal -> Interval -> Ratio.
Nominal = ?
Category/identity.
Ordinal = ?
Rank/order.
Interval = ?
Order + equal intervals + arbitrary zero.
Ratio = ?
Order + equal intervals + true zero.
load()
Opens an R-formatted dataset.
names()
Prints the names of all variables in a dataset.
head()
Prints the first few rows of a dataset.
str()
Reveals key structural information about variables, including variable type.
attr()
Provides information about an object's attributes and can help interpret the meaning of coded numerical values.
table()
Provides frequency counts for each level or value of a variable.
mean()
Computes the arithmetic average of a set of numerical values.
factor()
Reclassifies a categorical variable so its values are treated as categories rather than quantitative numbers.
Steps for reclassifying a numerical categorical variable as a factor
Use factor() to reclassify and label categories, run str() to confirm, run table() to confirm frequencies, and check mean() returns NA.
What are the 3 building blocks of psychological measurement?
Differences/variability, Consistency/covariability, Interpretation of test scores.
Variability
The degree of differences within a set of test scores or among values of a psychological attribute.
Covariability
The degree to which variability in one set of scores corresponds with or is consistent with variability in another set.
What assumption does psychological measurement rest on?
People differ in behaviour and psychological characteristics, and those differences can be detected and quantified.
Interindividual variability
Differences that exist between people.
Intraindividual variability
Differences that emerge within the same person over time or under different circumstances.
Interindividual vs. intraindividual
Interindividual = between people. Intraindividual = within one person over time/circumstances.
Distribution of scores
A set of test scores obtained from a group of people or from different measurement occasions.
Variance
A statistical way of quantifying variability or individual differences in a distribution of scores.
Covariance
A statistic quantifying the connection between variability in one distribution and variability in another distribution.
Variance vs. covariance
Variance = variability within one variable. Covariance = how variability in two variables is connected.
Central tendency
The score that is most representative of an entire distribution.
Mean
The average or "typical" score in a distribution.
X
An individual's score.
Σ
An instruction to sum the values that follow it.
N
The total number of people, observations, or X values in a distribution.
What are the 2 main statistics used to quantify variability?
Variance and standard deviation.
What do variance and standard deviation represent?
The degree to which scores in a distribution deviate from the mean.
What are the 3 steps for computing variance?
Calculate deviation from the mean, square each deviation, compute the mean of the squared deviations.
Deviation score
The difference between an individual's score and the mean, indicating how far that score lies above or below the mean.
Sum of squares
The sum of all squared deviations around the mean.
Variance conceptually
The mean of the squared deviation scores.
Standard deviation conceptually
A measure of variability reflecting the size of the raw deviations from the mean.
Why is standard deviation often more intuitive than variance?
Standard deviation reflects variability in the raw score metric, whereas variance reflects squared deviations.