Psychometrics: discipline dedicated to construction & validation of standardized instruments
Psychometric Properties: characteristics of an assessment tool that reveal info about the validity & reliability of a tool and provide evidence of the measurement quality
Construct: a concept, skill, or ability that is being assessed by the test
Correlation: level of association between two variables but DOES NOT imply causality but identifies that a relationship exists
Correlation Coefficient (r): statistic that indicates the degree of relationship between two measures
higher the (r), the greater the strength of correlation increases, greater the relationship
.90-.99 = standard - very high & preferred but not frequently observed (eg. chance, certainty)
.80-.89 = satisfactory or adequate
Significance (p): aka - statistical significance or probability; identifies the likelihood that a particular outcome may have occurred by chance
the smaller the p-value, the greater the likelihood the findings are valid
p=.05 or smaller - standard
Reliability: the degree to which a measure yields consistent results (does the test consistently provide a similar score with each administration)
Test-Retest: measures test score's stability over 2 testing occasions under same conditions
Intrarater: stability of scores on a instrument when using the same rater across repeated measurements
Interrater: stability of scores across different raters
Internal Consistency: the degree of agreement between items in a measure that measures a single construct
Validity: degree to which a variable measures what we think it is measuring (does the association test what it is supposed to?)
Face Validity: the appearance of test items in relationship to the purpose of the test
Content Validity: extent to which the test items represent the construct being measured
Criterion-related Evidence of Validity: establishes that the results/scores from one measure match those obtained with a more direct or already validated measure of the same phenomenon (the criterion)
Concurrent Validity: correlation of scores on an instrument with certain other measures that may be similar to the construct or closely associated with the construct
Predictive Validity: ability of scores on the measure to predict future abilities or outcomes
Construct-related Evidence of Validity: relationship between the underlying theory of the construct and the scores of an instrument
Convergent Validity: correlation of scores on an instrument with certain other measures that may be similar to the construct or closely associated with the construct
Divergent Validity: scores on an instrument diverge from those of other measures with a different construct & demonstrate a low correlation
Responsiveness: the ability of the scores on a measure to indicate change as a result of an intervention
Minimal Detectable Change: changes in scores beyond the measurement error of the instrument
Minimal Clinically Important Difference: changes in scores that result in substantial gains in the person that are meaningful
Floor Effect: when the scores on a measure are not able to detect a change at the lower level of ability
Ceiling Effect: observed when the scores are not able to detect a change in the higher ability level
Sensitivity: ability of scores to correctly detect a condition or impairment with high accuracy
Specificity: ability of the test scores to correctly reject those people who do not have the condition or impairment
Ratio: true zero point, usually used for physical measures (eg. miles per hour, ROM, height, weight)
Interval: equal intervals among measurements (eg. temp, timed task)
Ordinal: ranks the characteristics along a continuum (eg. high to low stress, tallest to shortest, MMT, Likert scale)
Nominal: having or not having a characteristic; names or categories (eg. female or male, present or not present, yes/no questions)
Measure of Central Tendency: a summary measure that describes a set of data with a single value that represents the middle of its distribution
Mean: average value of data set (most preferred)
Median: value in the middle of the data set (second most used)
Mode: most frequently occurring score in a data set (least used)
Normal Curve: a specific type of bell-shaped curve that displays the normal distribution in which most of the scores fall in the middle of the scale and fewer fall at the ends
Variability: how far or closely together data points are from each other and from a measure of central tendency
Standard Deviation (SD): how far or close scores/data points are from the mean
Variance: how far or close a score/data point is from the mean
Range: spread of scores/data points from lowest to highest scores
Interquartile Range: measures the spread of the middle half of the data; range for 50% of the data