Foundations of Quantitative Measurement Flashcards
What is quantitative research ?
All variables can be converted to numeric form and analyzed using statistics. Quantitative–> quantity
What is qualitative research?
More exploratory and focuses on non-numerical data
Classical Test Theory
Observed score = true score + error. Can underpin the concept of reliability and validity
Psychological Construct
A psychological concept that is not directly observable. It is not an object in the world;it does not have tangible existence outside of a person’s mind. (e.g. anger)
Operational definition
A clear, measurable definition of a construct based on a theory. It may capture only a portion of the entire construct.
“Anger –> a specific constellation of facial muscle movement cross-culturally found”
Measure
A way of observing or testing a construct.
Depression - Beck Depression Inventory Score
Motivation - number of button presses to receive reinforcement
What is Validity –> Construct Validity (UMBRELLA TERM FOR ALL VALIDITY)
Construct validity refers to how well you translated or transformed a concept, idea, or behaviour - that is a construct - into a functioning and operating reality, the operationalization
Content Validity
Assesses whether the measure adequately covers the different aspects of the construct that are specified in its definition. E.g. does a self-report measure of depression scale have items that capture the components of lowered mood, decreased motivation, etc.?
Face Validity
Assesses whether the measure looks right on the face of it, that it self-evidently measures what it claims to measure.
The extent to which a measure “appears” to measure the underlying construct.
E.g. some personality disordered symptoms may have poor face validity due to the developmental changes in their lives.
Criterion Validity and 2 subtypes
How well does the measure correlate with established “gold standard” measures of the same construct.
Concurrent validity: “At the same time –> correlate questionnaire with a clinical interview”
Predictive validity: “Predicting the future, e.g. does your measure predict future weight loss?”
Sensitivity
How well a measure picks out someone who actually has the disorder (few false negatives)
Specificity
How well a measure avoids diagnosing healthy people with a disorder (few false positives)
ROC Curve
LARGER values on the y axis indicate better sensitivity.
SMALLER values on the x axis indicate better specificity
Convergent Validity
How well the construct correlates with other measures that it should be related to (e.g. self report and clinical diagnosis of an eating disorder)
Discriminant Validity
How well does a measure not correlate with measures that it should not be related to (e.g. schizophrenia symptoms and eating disorder should not be strongly related)
Reliability
Refers to the degree of reproducibility of the measurement.
E.g. if you repeat the measurement in various ways, do you get the same results each time?
Test-retest reliability
The measure is administered over two or more separate occasions among participants (key in longitudinal studies). More appropriate to assess personality over time compared to fear.
Internal Consistency
Assessing the inter0item reliability of a sale that is composed of multiple similar items. Used to assess a questionnaire with multiple items. Do all of the items in the questionnaire measure the same thing?
Chronbach’s Alpha
Used to assess internal consistency among items in a given measurement. In order to increase a, one must either increase the items of measurements, or remove items that are weakly correlated with the other items.
Inter-rater reliability
It is the extent to which coders’ ratings agree or vary with each other.
Used for observation rather than self-report measures in order to check the reliability of observations.
Nominal scales: Inter-rater reliability
% agreement of rating which were identical across ratings.
Can also use ordinal, ratio, or interval scales to test this.
How to know if a construct/measure is reliable or valid based on statistics?
Reliability: 0.80 (good), 0.50 (poor)
Validity: 0.50 (good), poor (0.10)
Descriptive Hypothesis
What is X like?
“Schizophrenia will occur very rarely in the general population”
Descriptive/Comparison Hypothesis
Does group 1 differ from group 2?
“Alcoholics with have more relationship problems than non-alcoholics”
Correlation Hypothesis
Do X and Y covary?
Psychopathy will be positively correlated with criminal behaviour
Psychometric Hypothesis
Is a measurement reliable and valid?
The Beck Depression scale will be a reliable and valid measure of depression
3 Most Common Hypothesis in Published Psychology Research
Mediation: X leads to M (mediator) which in turn leads to Y. (Perfectionism leads to increases in interpersonal conflict (mediator), which in turn predicts depression (DV).
Moderation: The relationship between X and Y varies depending on the value of the moderator, M (e.g. the relationship between stress (IV) and depression (DV) depends on an existing vulnerability(M). Specifically, stress only leads to depression in the presence of low social support)
Incremental validity: x1 predicts Y over and above another known predictor x2 (e.g. Enhancement motives will predict future binge drinking over and above past drinking)
R2
The percentage that the circle overlaps is what this statistic refers to (accounting for frequency or variability)
