Study and sheet notes

Het eerste deel van de vragen gaat over het artikel van Firk et al. (2018) met als referentie:

Firk, C., Konrad, K, Herpertz-Dahlmann, B., Scharke, W., & Dahmen, B. (2018). Cognitive development in children of adolescent mothers: The impact of socioeconomic risk and maternal sensitivity. Infant Behavior and Development, 50, 238-246. https://doi.org/10.1016/j.infbeh.2018.02.002

1. What is the research question of Firk et al. (2018)?

What are the differences in children’s cognitive development between children of adolescent and adult mothers in their first two years of life? Are these differences between children of adolescent and adult mothers mediated by maternal parenting qualities?

2. What is the hypothes is of the researchers with regard to this research question?

No clear hypothesis is mentioned, but Firk et al. do repeat what they want to investigate: They want to test the hypothesis whether maternal socioeconomic problems have a direct effect on child outcome or whether this effect is mediated by maternal sensitivity.

3. Describe in your own words which groups the researchers want to compare. How did the researchers select these samples?

Mother-child dyads were recruited within a project on adolescent parenting in cooperation with the local youth welfare system, obstetric clinics, midwife practices and pediatrician practices. Adolescent mothers were all < 21 years, adult mothers were all > 25 years. It is unclear how the samples were actually selected, but it appears that all dyads that wanted to participate and met the inclusion criteria, were included in the study.

4. In the results section, study the first mediation model that the researchers analyze. What is in this model the independent variable (X), the dependent variable (Y) and the mediator (M)?

X = Maternal age
Y = Cognitive development
M = Maternal sensitivity

5. For model 1, use the values included in the text and in the notes of Figure 2, to fill in the table below. For each effect, note (a) the lower and upper boundaries of the confidence interval, (b) whether the value 0 falls between these boundaries, and (c) whether this effect is significant.

6. What is the conclusion of Firk et al. (2018)? Is this in line with their hypothesis?

When controlling for maternal IQ, children of adolescent mothers showed lower cognitive skills than children of adult mothers. In the current study, this difference emerged when children were between 18 and 24 months of age. In contrast to children of adult mothers, whose age-standardized scores were comparable at all time-points, age-standardized scores for children of adolescent mothers decreased from T2 to T3.

When controlling for maternal IQ, Firk et al. found that the effect of adolescent parenthood on cognitive development was to a great part explained by maternal sensitivity. A serial mediation of socioeconomic risk and maternal sensitivity demonstrated that adolescent motherhood was associated with socioeconomic problems that affected maternal sensitivity, which in turn affected child cognitive development.

Even though the authors do not mention a clear hypothesis, most of the results appear to be in line with their expectations and results from previous research. However, they do state that contrary to their expectations, socioeconomic risk did not mediate the effect of adolescent parenthood on cognitive development in a simple mediator model.

De volgende vragen gaan over statistische concepten. 

1. Explain in your own words what between- and within-subjects-factors are and give an example. 

Within-subjects-factor = factor that separates multiple measures within the same individual. For example, the factor time: Knowledge of statistics measured before, during and after ARMS.

Between-subjects-factor = factor that separates groups of different individuals. For example, the factor group: Knowledge of statistics measured in students of the four different studypaths of ARMS.

2. What would you call ananalysis of variance that only includes within-subject- factors?

Repeated measures ANOVA

3. What would you call an analysis of variance that only includes between- subject-factors?

(Factorial) ANOVA

4. What would you call ananalysis of variance that includes within- as well as between- subject-factors?

Mixed design ANOVA

5. Explain each of the concepts below in your own words.

Total effect = The simple relationship between the independent variable (X) and the dependent variable (Y) is shown in the figure below. In mediation, this effect is often called the total effect (c).

Direct effect = The direct effect is the effect of the independent variable X on the dependent variable Y, controlled for the mediator M (c’).

Indirect effect = The indirect effect is the effect of the independent variable X on the dependent variable Y through the mediator M. This effect is obtained by multiplying path a and b.

Complete mediation = With complete mediation, the direct effect of X on Y disappears completely when the mediator M is added to the model (c’ = 0). In that case, only an indirect effect is present.

Partial mediation = With partial mediation, a part of the direct effect of X on Y disappears when the mediator M is added to the model. In that case, both a direct as well as an indirect effect are present in the model.