Chapter 3 Flashcards
Variables
characteristics and conditions that change of have different values for different people, animals or objects.
Measurement
a process through which the nature, quantity or intensity of something is determined.
process of measurement
1) define a construct
2) measurement instrumentation
theory
why something is occurring, has occurred, or will occur
contructs
variables that are hypothetical factors, attributes, or mechanisms that cannot be observed DIRECTLY, but are inferred from some behaviour.
attributes of mechanisms that help explain and predict behaviour in a theory. You can study the behaviours that theoretically result from a construct.
Drawbacks of using operational definitions to measure the construct
not the same as the construct itself. an operational definition is only explained as observable traits, but a construct might be able to be defined via multiple traits, including cognitive and emotional aspects (for the construct of lonliness, introversion may be an emotional trait that cannot be operationally defined because it is not observable.)
how can you be sure that measurements obtained from an operational definition actually represents the construct?
by measuring validity
What is construct validity?
determines whether the measurements of a variable behave in exactly the same way as the variable itself, or whether the tool is the best for the construct, and accurately assesses it. How well does the measurement assess the construct?
Three types of construct validity
1) face validity; does it make sense?
2) content validity: degree to which the measured variable appears to have adequately samples frrom the potential domainds of questions that might relate to the construct of interest.
3) convergent and divergent validity.
convergent= creating 2 different methods to measure the SAME construct, then showing a strong relationship between values obtained from the 2 measures.
divergent= demonstrating that the measurements on one specific construct does not include a different construct in the measure.
criterion-related validity
validity used by some standard or criteria THAT IS KNOWN in order to indicate a construct accurately. verified by comparing with another measure of the same construct in which the research has confidence.
example of criterion-related validity
the SAT is valid to the extent that it distinguishes between students that do well in college versus those that do not.
two types of criterion-related validity
1) Concurrent Validity; a new measurement technique that is directly related to scores from another, better establish technique.
2) Predictive Validity; when measurements of a construct accurately predict behaviour according to a THEORY
Reliability
a measurement procedure produces identical results when it is used repeatedly.
a measure is reliable if there is very little ______
measurement error, or little deviation
Name types of error that may effect reliability
1) observer error: person tracking the observations may be distracted, the equipment used to make measurements in the experience may be compromised
2) environmental changes: time of day, room temp chaned etc
3) participant changes: stress, transient mood, health, fatigue level, may change outcomes.
ways of measuring reliability?
1) test retest
2) inter-observer reliability; analyzes degree of agreement between 2 or more observers that simultaneously record measurements independently
3) internal consistancy: the extent to which scores on the test correlate with each other
What is split-half reliability
split-half reliability compares the results of one half of the test with the results of the other hald, aids in the measurement of the extent to which all parts of the test contribute equally to what is being measured.
T/F a measurement can be reliable but not valid
true
Nominal scale of measurement
Qualitative/categorical data such as names or classfication (sex) that cannot be ordered whatsoever. mutually exclusive and exhaustive.
ordinal scale of measurement
ordered or ranking of categories
-data can be ordered sequentially, but differences between categories cannot be determined
example of ordinal data
shirt size, grades.
interval data scale of measurement
numeric and quantitative with an arbitrary zero point. categories must all be the SAME SIZE
example of interval data
temperature.
Ratio scale of measurement
numeric or quantitative with a meaningful 0 point, which indicates the complete absense of a variable being measured- data can be compared in terms of ratios
example of ratio data
weight (comparison–> he is 4x (ratio) heavier than other person)
Modalities of measurement
1) self report measures
2) physiological measures
3) behavioral measures
pros and cons of self report measures`
Pros: most direct way to assess a construct, more apparently valid.
cons: easy to distort values to make the individual look better,
hello good bye effect: patients tend to exaggerate their symptoms at the beginning and minimize symptoms at the end
pros and cons of using physiological measures as a modality of measurement
advantages: objective, utilizes accurate and well defined measures.
disadvantages: expensive equipment, the physiological measure may not accurately measure the construct.
pros and cons of using behavioral measures as a modality of measurement
advantage: many options of behavior to use- can be natural or task basked measures. can tailor to individual patients. the behavior itself may be the variable of interest.
cons: behaviour may be temporary or situational, observer can be subjective in what they see. ex/ disruptive vs playful
pros and cons of using multiple measures to measure a construct
multiple measures: using 2 or more different modalities of measurement to analyze the same variable
pros: provides more confidence in validity
cons: requires complex multivariate statistical analysis, variables may not behave the same way- one may be more sensitive than the other, causing different dimensions of the variable to change at different times
desynchrony
when 2 variables that are being used as multiple measures do not agree between each other.
examples of desynchrony
studying therapy dos aids in anxious situations by measuring heartrate and behavior
-therapy dog exposure results in a large effect on behavior, but NO EFFECT on HR; HR may still race in an anxious situation, even if the person appears composed behavior)
ceiling effect indicates _____
insensitivity- measurement range is restricted at the high end/ ceiling.
example of ceiling effect
giving a reading test meant for four year olds to 12 year olds- probably every subject would score in the 95th percentile
floor effect indicates_____
too sensitive- measurement range is restricted at the low end of the spectrum
example of floor effect
using an analytical chemistry exam as a measure of intelligence in 5 year olds- theyd probs all score poorly
range effects indicate that there is _____ between modality of measurement and the individuals chosen to be measured
incompatilbility
what is an artifact? example?
any EXTERNAL factor that can affect measurements- a threat to validity and reliability
ex/ using an ice cold stethoscope to measure HR- the cold (external factor) may startle the patient, increasing their HR. This HR reading thus may be incorrect (invalid) and may be different if measured again when the stethschope had a chance to warm up to room temp (low reliability)
2 notable artifcats and how to prevent artifacts
1) experimenter bias- use double blinded studies
2) demand characteristics and participant reactivity- any potential cue of a study that suggest to participants what the purpose or hypothesis is of the study, which may influence the participants to behave a certain way
any potential cue of a study that suggest to participants what the purpose or hypothesis is of the study, which may influence the participants to behave a certain way
demand characteristics- type of artifact that may be a threat to validity and reliability
Different Types of subjects
1) “ good “subject- has identified the hypothesis and acts a certain way to support the investigators hypothesis
2) “negative” subject- has identified the hypothesis and acts a certain way to sabotage the results
3) apprehensive subjects- individuals try to place themselves in a desirable light by responding in a socially-acceptable fashion
4) faithful subjects- answer honestly, ideal subject.
strategies to combat reactivity
1) use a field study instead of laboratory setting
2) reassure participants that their performance and responses are confidential
