Term
|
Definition
| definition of a variable in terms of the activities/variables/operations the researcher uses to measure or manipulate |
|
|
Term
|
Definition
| variable you manipulate or change (manipulate the independent loner) |
|
|
Term
|
Definition
| the variable being studied/measured, (the control) |
|
|
Term
|
Definition
-categories, categorizing, not numbers separated into groups
-Gender (all IV's) |
|
|
Term
|
Definition
-scale in which objects or individuals are categorized by rank
-places in a race, class rankings |
|
|
Term
|
Definition
-numbers on the scale are all equal in size identity, magnitude and equal unit met
-Likert rating scales, most psychological scales |
|
|
Term
|
Definition
-Absolute Zero is possible
-height, weight, age, time |
|
|
Term
|
Definition
-Data measured at interval or ratio levels
-average of the distribution
-Not too much skew or outliers |
|
|
Term
| (Central tendency) Median |
|
Definition
-data measured on ordinal scale
-mean is not valid
-the middle score of distribution after scores have been arranged lowest to highest or vice versa |
|
|
Term
|
Definition
-Nominal Data
-score that occurs with the greatest frequency |
|
|
Term
|
Definition
| -Middle of whole distribution |
|
|
Term
|
Definition
-the extent to which the results can be generalized
-Diverse sample increases it |
|
|
Term
|
Definition
-the peak is to the left and the tail is to the right
-Report Median, Minimum, and Maximum |
|
|
Term
|
Definition
-the peak is to the right and the tail is to the left
-Report Median, Minimum, and Maximum |
|
|
Term
|
Definition
-Report M and SD
-Normal curve: symmetrical, bell-shaped frequency representing normal distribution
-(mean, median, mode are equal in the center, only one mode, observations clustered in center) |
|
|
Term
| (Types of Design) Correlation |
|
Definition
-(Pearson Correlation, must be interval or ratio data)
-Any relation, no manipulation |
|
|
Term
| (Type of Design) Descriptive |
|
Definition
| -Gathering data ex. survey |
|
|
Term
| (Type of Design) Experimental (Simple) |
|
Definition
-1 IV with 2 levels
-run independent samples t test
-The assumptions of the independent samples t test are that the IV is nominal and dichotomous (2 levels)
-The DV is always interval or ratio
-The levels are independent from each other
-Advantage – Random assignment
-Limitation – Confounds, other things playing an effect other then the IV |
|
|
Term
|
Definition
| Rejection of null hypothesis. p<.05 |
|
|
Term
|
Definition
| Retain null hypothesis. p>.05 |
|
|
Term
|
Definition
| What you want to reject. (key words: “no diff”, “not) |
|
|
Term
| Alternate Hypothesis (Directional & Non-Directional) |
|
Definition
-Predicts what you hope to find
-Directional Hypothesis:
-predict direction “_____will increase_____”
-Non-Directional:
-predict a difference (more general) "_____will
have a relationship with______" |
|
|
Term
|
Definition
-Outside influences (outside the study)
-Sometimes use MATCH design to get around confounding variables
-Ex. Being hungry, tired, angry, etc. |
|
|
Term
|
Definition
-Within the design of the study
-Pilot study – ask other if the study asks what needs to be asked
-Threats – not measuring what needs to be measured
Strong internal validity, the variables are reliable, measures are
-Ex. Use Crombach’s Alpha, want to be equal or above .7 |
|
|
Term
|
Definition
-Getting means and standard deviations
-See the differences within the stats, but don’t know if the difference is significantly significant
-Analyze – Descriptive Frequencies |
|
|
Term
|
Definition
-Can now get the information about significant significance
-Correlation, t test, ANOVA
-Gives p value (Sig.) |
|
|
Term
| Increase Power (Effect Size) |
|
Definition
-Ability to detect a difference in the DV based on the manipulation of the IV
-Increase:
Get rid of confounds
Strong manipulation of the IVs, use more extremes
of the variables
Increase participants
Decrease within group means |
|
|
Term
| Between (or treatment) variance-difference between the groups |
|
Definition
|
|
Term
| Within groups (or error) variance – differences between individuals within one group |
|
Definition
-Want to be a low
-If you have a high number, this could mean that there are other variables (confounds) effecting the results |
|
|
Term
| Statistical Significance testing (including what impacts the likelihood of statistical significance and what p < .05 means) |
|
Definition
-Inferential Statistics
Correlation, t test, ANOVA
Gives p value (Sig.)
p < .05 – Significant difference between the groups
p > .05 – No Significant difference between the groups |
|
|
Term
| Effect size (proportion of variance accounted for), rpb2, eta2 |
|
Definition
–percentage of effect that is directly attributed to the IV manipulation
-Ex. If the rpb2 is .32 or 32% then it can be said that of everything that could effect math scores, 32% of the effect is directly due to the type of music the participant listened to. The other 68% percent of the effect on math score is due to other reasons. |
|
|
Term
|
Definition
-Difference between the groups, just because it is statistical significant it might not be practically significant, the effect size could be large or small
-p < .05 |
|
|
Term
|
Definition
-In the real World, Does it matter? Does it make a difference?
-Will have practical significance when there is statistical significance and a high value effect size. |
|
|
Term
|
Definition
-How much of effect that is directly attributed to the IV manipulation?
-Want to be large
-Ex. How much of the effect on math scores in directly due to manipulation of music |
|
|
Term
| (Type of Design) Experimental (multi-group multi-level) |
|
Definition
-1 IV with more than 2 levels
-run one way ANOVA, if significant difference run post hoc (LSD)
-Advantage – Random assignment
-Limitation – Confounds, other things playing an effect other then the IV |
|
|
Term
| (Type of Design) Experimental (Dependent) |
|
Definition
–Advantages – Less worries of confounds, more control over outside influences
-Match Design
Matched on a characteristic (age, score, gender,
etc.) with another, then paired, then doing a
dependent samples t test on that pair based on
-Limitation – Mortality/Attrition, if you us lose one participant in a group you lose the whole group (match)
-Repeated/Within Design
Participants receive all conditions in random order
-Limitations – Mortality/Attrition, participant may not come back for repeated study, study may take a lot of time to complete. |
|
|
