Help
Options
Didn't know it?
click below
click below
Knew it?
click below
click below
Don't Know
Remaining cards (0)
Know
retry
shuffle
restart
0:00
Embed Code - If you would like this activity on your web page, copy the script below and paste it into your web page.
Normal Size Small Size show me how
Normal Size Small Size show me how
PSYC 204- ASSIGN2
In class assignment 2
| Term | Definition |
|---|---|
| Nominal Scale | -Data can be categorized --Numbers represent categories --No overlapping values |
| Ordinal Scale | -Data can be categorized, ranked (relative to other categories) --Cannot be used to derive statistics like arithmetic mean --Likert Scales |
| Interval Scale | -Data can be categorized, ranked, evenly spaced/equal values between intervals (no ambiguity) --No true zero |
| Ratio Scale | -Data can be categorized, ranked, evenly spaced/equal values between intervals, and has a natural zero. --Allows make of proportional statements about measurement data ---Only proportional statements about non-zero numerical values ---Temperature |
| Two rules of nominal scale | -Mutual exclusion -- A "case" (what you're measuring) can only fit into one category -Exhaustivity --Include a residual category (other/unsure) |
| Nominal scale: Predefined Response Categories vs. Specify Responses | -Other -Other please specify: ______ --Provides info that researchers might not have anticipated when designing a measure |
| Residual categories | -Attempt to capture missing data -- Other/unsure/specify___ |
| Direct vs. Indirect Indicators | Direct: -Indicator is an indicator of another thing Indirect: -Sherlock logic |
| Multiple Indicator Measures (MIM) | - For abstract or complex variable measurement (ex. intelligence) -Measure dimensions of a variable (ex. subtypes) |
| Categorical data | Nominal or ordinal scale |
| Continuous data | Interval or ratio scale |
| Reliability | Consistency in results |
| Validity | If the measure actually measures what it's designed to -All or nothing (ex. yes or no) |
| Reliability Test: Test-Retest Reliability | Test something using a measurement instrument and then test the same thing again to see if the result is same |
| Reliability Test: Equivalent Forms | -When equivalent tests exist, researchers build a question bank and use these to create multiple versions of a test --Rewording questions on the test/measure |
| Stats: Response Similarity | -Item by item comparisons of responses --Mode (with categorical measures) OR -Aggregate score comparisons --Mean comparison or correlational analysis (interval or ratio level data) |
| Stats: Response Similarity: BEWARE | -Altering precisely worded questions may alter the meaning/responses --Underestimating a statement change's affect on the degree of reliability |
| Reliability Test: Internal Consistency Reliability | -Used in MIM when indicators measure the same thing --Using Split-Half method |
| Split-Half Method | -Internal consistency reliability -Split indicator data into two sub-groups, then compare results using correlational analysis |
| Stats: Split-Half Method: BEWARE | -Some test items will measure different dimensions of a variable --Low correlation may make a measure seem less reliable than it actually is |
| Reliability Test: Inter Rater Reliability | -Based on level of agreement between two or more independent observers (who are studying the same thing) --Quality consistency |
| Pseudo-scientific theories | -Based on reliable but invalid measurements |
| Accuracy | -Degree of truth |
| Evidence for Validity (based on content): Face Validity | -Extent to which a measure/indicator seems like it measures what it claims to measure |
| Face Validity: BEWARE | -Easier for test taker to "game" tests (produce an outcome they want) --Reduces reliability |
| Fix: Face Validity | -Use indirect measures with low face validity (harder for test taker to interpret) --Less obvious answer ---Lower chance of participants being able to "game" the test |
| Evidence for Validity (based on content): Psychometrics | -Specialized in measurement and testing --Determine if someone is actively trying to "game" the test/isn't paying attention ^By comparing high face validity and low face validity/indirect question responses: ---similar=unlikely ---dissimilar=likely |
| To "game" a test | -Knowingly or unknowingly manipulating results of a test to achieve a desired outcome |
| Evidence for Validity (based on content): Criterion Validity | -How well a measure compares to another, well established measure |
| Criterion Validity: BEWARE | -Be sure the comparative measures are valid too --Investigate sources of "valid" information! |
| Evidence for Validity (based on content): Predictive Validity | -Predictive accuracy based on present data (or "scores") --Predict an outcome based on correlational data |
| Statistic | -An ESTIMATE of the parameters of a population --Stats are always/only an estimate |
| Parameters | -Actual characteristics of a population |
| Error | -Difference between sample statistic and population parameter --"Luck of the draw" phenomenon ---Sample population "luck of the draw" type condition ----May not represent general population |
| 2 Ways to Reduce Errors (caused by sampling) | 1. Increase sample size 2. Improve representativeness -Random sampling |
| Random Sampling | -Each member of a population has equal chance to be chosen for participation --EPSEM rule (equal probability of selection method) |
| Sampling Procedures(3): Identify Target Population | 1. Identify target population 2. Element list -Eligibility criteria --Pool from which the sample is drawn 3. Decide on sample type |
| Purpose of Samples | -Samples are used to make generalizations about populations |
| Random Sampling: BEWARE | - Obtaining diverse participants for research --Response rates differ among different groups of people |
| Types of Strategic Sampling | -Cluster sampling -Stratified sampling |
| Cluster Sampling | -Split an element list into several sub-categories and use census data to find population percentages, then randomly sample participants in their matching categories from the element list |
| Strengths and Weaknesses: Cluster Sampling | Strength: -Improves representativeness -Preserves EPSEM principle Weakness: -Sampling error is greater because each sub-category contains smaller participant samples |
| Stratified Sampling |
Created by:
user-1982862
Popular Psychology sets