General Linear Model Flashcards
interpretation of the basic General Linear Model
- y(i) = score on the DV for subject i
- x(i) = intervention (0 or 1) used for subject i
- b0 = the intercept = the value of y when x is 0 -> the average [DV] for [IV=0] -> average depression at posttest of the sitting group.
- b1 = the slope = the change in y when x goes up by 1 -> the difference in average [DV] between [IV=1] and [IV=0].
- error(i) = the difference between the observed and then expected DV score for subject i
3 tests that are based on the General Linear Model
- regression analysis
- t-test
- one-way ANOVA
- factorial ANOVA
wat test je met een regressieanalyse
test whether x can predict y, if y is a continuous variable
hoe doe je een regressieanalyse in SPSS
- Analyze -> regression -> linear -> place DV in dependent -> place IV in independent(s).
- Statistics -> estimates | Plots -> histogram and normal probability plot | Save -> standardized residuals
waar kijk je naar & wat interpreteer je bij regressie analyse
-> When F is significant x significantly predicts y.
kijken naar: ANOVA
* [IV] [significantly predicts/does not significantly predict] [DV], F([df], [df error]) = [F], p = [Sig.].
* On average, the patients in the [control group] report [interpretation of the DV] at [measurement moment], t([df]) = [t (Constant)], p = [Sig.(constant)].
kijken naar: Coefficients
* On average, [dependent variable] at [measurement moment] are [lower/higher] ([b1] points) for patients in the [condition 1] than in the [condition 2], t([df]) = [t(b1)], p = [sig.(b1)].
*Treatment significantly predicts social anxiety
complaints at posttest, F(1, 138) = 7.78, p = .006.
*On average, patients in the waiting list group
report social anxiety complaints at posttest,
t(138) = 70.14, p < .001.
*On average, social anxiety complaints at
posttest are lower (4.17 points) for patients in
the mCGT group than in the waiting list group,
t(138) = -2.79, p = .006.
waar kijk je naar voor de df
df 1 = regression
df 2 = residual
wanneer gebruik je een regressieanalyse
continuous DV
twee t tests en wanneer gebruik je ze
- paired samples t-test: when the conditions contain the same test-subjects
- independent t-test when they contain different subjects
hoe doe je independent t-test in SPSS
Analyze -> compare means -> independent-sample t test.
waar kijken & wat rapporteren bij t-test
-> When t is significant, the means are significantly different.
kijken naar group statistics:
* On average, the [condition] group had a [higher/lower] score on [dependent variable] (M = [Mean], SE = [Std. Deviation]) than the [control condition] (M = [Mean], SE = [Std. Deviation]).
kijken naar independent samples test:
* This difference ([mean difference]) was [not significant/significant], t([df]) = [t], p = [Sig.].
- On average, the mCGT group had lower
scores on social anxiety (M = 64.47, SE = 1.19) than the waiting list group (M = 68.64, SE = 0.94). - This difference (4.17) was significant, t(138) =
2.79, p = .006.
dependent t-test in SPSS
Analyze -> compare means -> paired-samples t test
t-test =
the difference between the means of the groups (X1 – X2) divided by the standard
error (SE). After calculating the value, we look it up under the null-hypothesis (H0)
distribution. If the t-value is in the critical area of the distribution, you conclude that
there are significant differences between the groups.