Term
|
Definition
-Commonly known as dependent variable
-Response measure, that value of which is depndent on the independent variables
-measure that captures the hypothesized differences |
|
|
Term
|
Definition
-Commonly known as independent variable.
-Variable that is manipulated
-Quantative, Qualitative, Classification
|
|
|
Term
|
Definition
- Created by selecting subjects ona dimension to be included in the study
- E.g sex of the subject |
|
|
Term
|
Definition
| -if left uncontrolled could exert a systematic influence on the different treatment groups |
|
|
Term
| Completely Randomized design |
|
Definition
| -subjects are randomly assigned to serve in one of the treatment groups |
|
|
Term
|
Definition
-Blocks of subjects who are matched on some relevant charcteristic
-Once Blocked, subjects within the blocks are assigned randomly to treatment consditions |
|
|
Term
|
Definition
- Randomized factorial design
and
-Block factorial design |
|
|
Term
|
Definition
-Defined by a set of membership rules
-Samples are a set from the population |
|
|
Term
|
Definition
- If parameters observed are too deviant from those specified by null hypothesis
-Usually states that the treatmnent populations are not all equal |
|
|
Term
|
Definition
- This represnts the popluation mean
-1st defining parameter of distrubution |
|
|
Term
|
Definition
-Represents the population standard deviation
-2nd parameter of distrubution |
|
|
Term
| Shared area of two distrubutions is small.. |
|
Definition
Then there is a large probability that the populations are different
-If the shared area is large then there is only a very small probability of the populations being different.
-Impossible for the shared area to be zero |
|
|
Term
| Assumptions of parametric statistical tests |
|
Definition
-ALl populations have normal distrubution
-Normal distrubutions have the same standard deviation (variance)
-Data from ratio or interval scale |
|
|
Term
| Individual differences error |
|
Definition
| -Come about because the nature of the individual subjects |
|
|
Term
|
Definition
| -Arises because of problmes with accurately collecting data |
|
|
Term
|
Definition
| -Estimate of variability due to particular treatmment condition |
|
|
Term
| Between group variability |
|
Definition
| -Estimate of variance due to differences between groups |
|
|
Term
|
Definition
| -Systematic source of variability |
|
|
Term
|
Definition
Treatment effects+ Experimental error
-------------------------------------
Experimental error |
|
|
Term
|
Definition
Experimental error
-----------------
Experimental error |
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
| -Subdividing of the total deviation, in to between groups and within groups |
|
|
Term
|
Definition
Sum of squared deviations from the mean
-------------------------------------
Degrees of freedom |
|
|
Term
|
Definition
-approximately equal to the number of numbers in a set
-number of scores with independent info which enter in to the calculation of SS, number that are free to vary with some knowledge |
|
|
Term
|
Definition
Differences among treatment means
---------------------------------
differences among subjects treated a like |
|
|
Term
|
Definition
For an overal difference between means
If significant requires further analysis |
|
|
Term
|
Definition
Planned comparrisons, express the difference between the means as the sum of weighted means
Degrees of freedom always 1 |
|
|
Term
|
Definition
- Used for quantitative data
-Use a weighted coeficent method of deciding whether or not the data describes a linear trend |
|
|
Term
|
Definition
1-β
Probability of not making a type 2 error |
|
|
Term
|
Definition
-sample size
-effect size
-Alpha level |
|
|
Term
|
Definition
| By making alpha less strict we can increase power, However we also increase type 1 error chance |
|
|
Term
|
Definition
-As the seperation between two means increase so does power
-As variability about a mean decrease, power increases |
|
|
Term
|
Definition
-Proportion of the total variance that is attributed to an effect
-Partial ''squared is proporption of the effect + error variance
-Mesure of association |
|
|
Term
|
Definition
-Proportion of variance explained by the model
-Mesure of association |
|
|
Term
|
Definition
-an estimate of the dependent variable populationvariability accounted to the independent variable
-Mesure of association |
|
|
Term
|
Definition
-Where there are only two groups it is the standardised difference between the groups
-Measure of differences |
|
|
Term
|
Definition
-Where there are 3 or more groups it acts as an averadged standardized difference
-small effect: 0.10, Medium effect: 0.25, Large effect: 0.40
-Measure of difference |
|
|
Term
|
Definition
-two independent variables
-one way, two way or three way |
|
|
Term
|
Definition
3 Independent variables
-One with 3 levels, one with 2 levels, one with 4 levels |
|
|
Term
| Mixed or Split plot design |
|
Definition
| Independent variables are a mix of between groups and within group variables |
|
|
Term
|
Definition
| Effect of the two independent variables together |
|
|
Term
| Analysing two way between group variables |
|
Definition
- State null hypothesis
-Partitio the variability
-Calculating mean squares
-Calculate the F ratios |
|
|
Term
|
Definition
-Homogenity of variance
-Between groups test |
|
|
Term
|
Definition
- Homogenity of variance
-Within Groups test |
|
|
Term
|
Definition
-HOmogenity of vairance
-Overall test, on mixed plot designs |
|
|
Term
|
Definition
-Used to gain statistical control of error when experimental control of error is not possible
-adjusts treatment effects with respect to the covariates |
|
|
Term
|
Definition
-Mainly collected before the experiment
-When not possible can be collected after |
|
|
Term
|
Definition
variance here has 3 basic components
-Effect
-Error
-Covariate |
|
|
Term
|
Definition
variance here has 3 basic components
-Effect
-Error
-Covariate |
|
|
Term
|
Definition
-Homogenity of regression coefficents
-assumption of linear regression. Cant use for curvelinear |
|
|
Term
|
Definition
| Similar distrubutions of scores according to the line of best fit |
|
|
Term
|
Definition
| Distrubution of scores according to line of best fit are not similar |
|
|
Term
| Homogenity of regression coefficents |
|
Definition
- Regression of coefficents for each of the groups in the IV are the same
- asses this by looking at the interaction between IV and covariable |
|
|
Term
|
Definition
- Only works best with a small number of covariables
-COvariates must be independent of treatmment |
|
|
Term
|
Definition
| Multivariate Analysis of Variance |
|
|
Term
|
Definition
| A new DV is created that attempts to maximise the differences between treatment groups |
|
|
Term
|
Definition
-Increses chance of finding an effect
-Protects from type 1 errors
--can be more powerful |
|
|
Term
|
Definition
-Homogenity of variacne-covariance
-Very sensitive so use a P value of 0.001 |
|
|
Term
|
Definition
if significant shows there is a multi variate effect
- if significant then need univariat tests |
|
|
Term
|
Definition
- fins a set of variable which predict group membership
-Discriminat functions analysis |
|
|
Term
| Discriminate Functions Analysis |
|
Definition
-Treats IV as the DV and reverse
-See if we can predict whos in which group |
|
|
Term
| Number of functions for DFA |
|
Definition
| - number of levels of the grouping variable - 1 |
|
|
