week one - intro to test and measurement Flashcards
definition of measurement
measurement is the act of identifying properties of an object
ex. carpenter would use a measuring tape to get the dimensions of a piece of wood
measurement theory
field of statistics that describes and evaluates the quality of measurements
examples of where we experience psychometric assessment
- testing in education to under stand student progress
- passing driving tests
- credit score testing
psychometrics - common set of fundamental characteristics
- they should all strive to be accurate
- they should measure what they intend to measure
- they should produce meaningful scores to make comparisons among individuals
- free of bias against members of certain groups
latent construct
seek to identify the properties of a psychological object
latent: can not observe it
construct - domain of behaviours
working memory example
goal: to access the working memory of two students
working memory is LATENT
Selected Test: Ability to recall strings of digits
Interpretation of Results: More digits recalled = more
working memory
individual vs groups tests
individual: given to one person at a time
group: given to a group of people simultaneously
achievement tests
assess prior learning
- persons degree of learning, success or accomplishment in a subject or task
aptitude tests
assess potential to learn
- capability for relatively specific task or skill
intelligence tests
ability to solve problems and think abstractly
personality test
measure traits, qualities, behaviours that determine person individuality
structured - accept or reject statements of ones self
(usually self report)
projective - reactions to ambitious stimuli are recorded and interpreted
(assumes responses reflect individual characteristics)
behavioural procedures
describe and count the frequency of a behaviour, identifying the antecedents and consequences of the behaviour
what is a test
technique to quantify a behaviour
composed of ITEMS
item: specific stimuli that produces a response
- usually require multiple items to properly assess the construct
psychometrics
the science of evaluating characteristics of a psychological test
criterion vs norm references tests
criterion referenced tests - decisions made in comparison to a cut off score
norm referenced tests - score is compared to some reference sample
- expected score from the population
- test taker needs to be part of reference population
Sir Francis Galton (1883)
came up with the term PSYCHOMETRY/PSYCHOMETRICS
- father of psychometrics
- suggests there are differences between individuals
- arguably the first to have a large focus on mental diff. bw individuals
- believes anything is measurable
Charles Darwin
describes individual differences in animals
Edward Jenner - 4 stage hierarchy of human intellect
highest to lowest:
1.mental perfection
2. mediocrity
3 being
4. idiots
who is the father of psychology as a science
Wundt
- creation of behaviour tests (ex. working memory)
The binet-simon scale (1905)
- first intelligence test to create
- test compared mental age with chronological age
- renamed to the stanford binet intelligence scale
army beta
specifically for those with little to no reading ability and used pictures and diagrams for assessing skills
world war 1
the testing was successful
- achievement tests began to be created and implemented
- by 1920s testing was occurring in nearly all industries and work forces
traits
relatively enduring dispositions that distinguish one person from another
Minnesota Multiphasic Personality
Inventory (MMPI)
- factor analysis made rapid developments
- allowing creation and assesmenent of complex constructs
factor analysis
allows the assessment of number of
dimensions (called factors) in a measure, identify a minimal set of items needed, and help confirm whether a measure
matches what we believe it is measuring (the factor
structure)
inferential statistics
provide predictions about a population based on data from a sample of that population
what are the 4 major variable types
ratio, interval, ordinal, nominal
Ratio Data
- continuous
- can rank values
- zero is meaningful
- can describe differences between values
ex. personal income, age, time
Interval Data
- continuous
- can rank values
- zero is NOT meaningful
- can only apply addition and subtraction
ex. Fahrenheit Temperature
Discrete Data
- values cant be fractions, only whole numbers
ex. number of children, count data
ordinal data
- categorical
- can rank/order the categories
- zero is NOT meaningful
ex. income categories, likert scales
Nominal data
- categorical
- can not rank/order the categories
- zero is NOT meaningful
ex. ethnicity, martial status, province of residence
Single-variable (Univariate) tables
left side indicates catgeories of the nominal or interval variable (rows)
types of info - frequency, percentage, count (collumns)
Relative Frequency
Number of times that value occurs/ number of observations in the data set
ex. suppose we have 200 observations on x = number of prior arrests. if 70 of these x values are 1 then
x = 70
70/200 = 0.35 (35%)
Mean
average from all the values in the sample
Median
represents the exact middle point in the sample
- in psychometrics, most likely to use a mean instead of median
Dispersion
variation among values
greater dispersion = greater range of scores = more possible differences could exist between scores
greater dispersion = central tendecy becomes less meaningful
Range
= maximum value - minimum value
Deviation
- measure of distance from the mean
Deviation = value - mean
standard deviation
square root of the variance
Variance
- Find the mean of the given data set. Calculate the average of a given set of values.
- Now subtract the mean from each value and square them.
- Find the average of these squared values, that will result in variance.
Percentiles
value where a specific percent of the remaining values will be less than it
- ex. if someone has an IQ test score in the top 5%, you can have evidence they are gifted
Correlation
measure of the behaviour between two variables
- strictly a measure of behaviour
Properties of Correlation
1.The value of r does not depend of which variable we
designate X or Y.
2.The value of r is independent of scale of X and Y
3.-1 <= r <= 1
4.r=1 if and only if all (xi,yi) pairs lie on a straight line
with positive slope. r=-1 in the case for a negative
slope
Spearmen Rho
correlation coefficient for ordinal data, commonly used for psychological measurement
- can not use for categorical data