STATISTIK-FORUM.de

[Leo4844]

Hallo!
muss für eine wissenschaftliche Arbeit Studien auswerten, aber kenne mich in dem Bereich leider nicht gut aus. Mich interessieren eigentlich nur die Effektstärken. Jetzt sind leider häufiger keine Variablen angegeben (bei pretest/ posttest) und ich bin mir unsicher, worauf die Werte sich beziehen. Sind das Mittelwerte? Hoffe jemand kann mir helfen. Vielen Dank schon einmal im Voraus!

Kann kein Foto anhängen, das ist der Text:
Intervention effects moderators
Mixed ANOVAs that differentiated participants according to certain criteria were conducted in order to evaluate if the effects of the programme can be obtained under certain conditions. The criteria included: degree to which the participants attended the workshops, sex, school level (elementary vs. high school), and aggression level at the pretest. Results evidenced significant interactions for self-report on anxiety and depression, as well as peer evaluation of aggression, prosocial behaviour, and anger management. First, we looked at differences in the effect of the intervention as a function of the time of exposure to the Yoga training. Given that mean attendance corresponded to 17 out of the 24 sessions, children who participated in the intervention were grouped into two categories: low-attendance (1 to 16 of the 24 sessions, n = 21) and high-attendance (17 or more of the 24 sessions, n = 47). t-Test comparisons indicated that low- and high-attendees differed significantly in their pretest levels of aggression (t (56) = –2.23, p < .05, low = .11, high = .15). Mixed ANOVA analyses showed an interaction effect of treatment exposition on empathy (Wilk’s Lambda .93; F (1, 66) = 4.82; p < .05) and depression (Wilk’s Lambda .87; F (1,66) = 9.89; p < .05). Contrast effects analyses showed significant differences in empathy between the low- and high-attendance group at posttest (low = 1.74, high = 2.31), but not at pretest. In the case of depression, contrast effects revealed significant differences between pre- and posttest for the high-attendance group (pretest = 1.33, posttest = .85), but not for the lowattendance group.

Comparing the effects of the intervention between elementary (n control = 32, n experimental = 42) and high school (n control = 25, n experimental = 26), mixed ANOVA results showed a significant interaction among treatment, school level, and repeated measures of depression (Wilk’s Lambda = .96; F (1, 121) = 3.95; p < .02). When the mean differences were examined closely (see Figure 1), contrast analyses revealed that pretest and posttest difference comparisons were found only for elementary school children. While depression increased for children in the control group (pretest = .73, posttest = 1.19), it decreased for children in the experimental group (pretest = 1.31, posttest = 1.03).

[PonderStibbons]

Das Messverfahren für Depression solte doch eigentlich irgendwo angegeben sein.
Dem Kontext nach ist anzunehmen, dass es sich um Mittelwerte handelt.

Mit freundlichen Grüßen

PonderStibbons