Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
27 Cards in this Set
- Front
- Back
Variable |
any characteristic that can tale on more than one value
|
|
Measurement |
a process by which we assign numbers to indicate the amount of some variable present
|
|
Four Levels of Measurement |
1. Nominal
2. Ordinal 3. Interval 4. Ratio |
|
Nominal Measurement |
Categorical data
no mathematical properties associated with this |
|
Ordinal Measurement |
Ranked Data
|
|
Interval Measurement |
Scored date/ continuous data
does not have a true 0 e.g., temperature |
|
Ratio Measurement |
Score Data HAS a true 0 e.g., reaction time |
|
Threat of Mismeasurement |
means by which we quantify our means of interest more precise our OD the more precise our mean of measurement and the more accurate the hypothesis |
|
Reliability |
the extent to which observed scores are free from errors of measurement consistent and reproducible |
|
Validity |
the extent to which the observed score reflects on the intended content are you measuring what you intend to measure? |
|
Psychometrics |
sub-discipline of psychology which focuses on judging and improving the reliability and validity of psychological measures
|
|
Observed Score |
1. True Score (Reflects the construct of interest)
2. Error (difference between the true and observed score) |
|
Random Error |
unpredictable should cancel each other out causes unreliability |
|
Systematic Error |
Predictable does not cancel out error that is due to reliability measuring the wrong construct not random error |
|
Bias |
Can arise from differences between measurement devices and differences between testing situations Solution: Standardized measurement and procedures |
|
Unreliability |
X= T + Ew + Er If Er=0, then X= T + Ew (reliable) |
|
Invalidity |
To be valid must be reliable reliability does not guarantee validity X=Ew (Invalid) |
|
Practical Issues |
all measures have imperfect reliability and validity all will contain some form of random error |
|
Estimating Reliability |
largest when there is no random error
smaller when entirely random error |
|
Three ways to Estimate Reliability |
1. Internal Consistency 2. Test-Retest reliability 3. Inter-rater reliability |
|
Internal consistency |
the extent to which scores on items measuring the same construct correlate with each other
alpha = 0, measure is entirely error alpha = 1, measure is error-free |
|
Test-Retest Reliability |
the extent to which the measure gives consistent results over time or across situations
|
|
Inter-Rater Reliability |
the extent to which the results are consistent across raters/judges
useful for more subjective measures |
|
Three Ways to Estimate Validity |
1. Criterion Validity
2. Content Validity 3. Convergent vs. Discriminant Validity |
|
Criterion Validity |
comparing the observations from our measurement device with some agreed upon gold standard |
|
Content Validity/ Face Validity |
put forward a case that your measurement device is tapping into all relative aspects of the construct interest |
|
Convergent vs. Discriminant Validity |
converge on the findings of measures of theoretically similar or related constructs and to diverge discriminate form measures of unrelated constructs |