Martine van der Pluijm

108 Chapter 4 TABLE 4.2: Means, standard deviations and effect sizes for Steps 1 and 3 in the pretest and posttest from the class inventory lists Pretest Posttest F(DF) p d Mean SD Mean SD Step 1: Gain insight into HLE • Insight in parental knowledge and skills • Insight in home language environment 4.26 3.87 .96 1.01 4.57 3.95 .91 1.07 4.259(33) 1.078(32) .05 .31 .33 .08 Step 3: Build reciprocal relationships 4.77 .71 4.90 .63 2.986(32) .09 .19 Measured on a 6-point scale: 1=disagree strongly, 2=disagree, 3=disagree a little bit, 4=agree a little bit, 5=agree, 6= agree strongly. We found no significant interaction effects of education level on change of the three constructs by within-subjects tests. We did find significant between-subjects effects of educational level on Insight in the HLE ( F (2, 32) = 15.629, p = < .000, d = .91), and on Build reciprocal relationships ( F (2, 32) = 7.562, p = .002, d = .61). Figures 4.4 and 4.5 show that the teachers’ ratings on these variables are generally lower at pretest for parents at Level 1 compared to ratings of parents with education Level 2 and 3. At posttest, ratings become more similar for the three education levels on the variable quality of reciprocal relationships (Level 1: M = 4.50 at post-test, M = 4.26 at pre-test. Level 2: M = 5.18 at post-test, M = 5.17 at pre-test and Level 3: M = 5.08 at post-test, M = 4.94 at pre-test). However, the ratings for both Reciprocal relationships and Insight in the HLE remain substantially lower at posttest for parents at Level 1, compared to the levels 2 and 3. These findings indicate that teachers indeed have less knowledge about the HLE and less access to parents with the lowest education levels compared to the groups of parents with higher levels of education. The questionnaire related to Step 2 [Teachers involve parents and colleagues in SFP procedures in support of child language development] includes questions to rate teachers’ ( n = 14) abilities to conduct the SFP procedures at pretest and posttest. Teachers reported significantly higher ratings of their abilities to plan their SFP procedures in support of child language development at posttest ( M = 4.89. SD =.63) compared to pretest ( M = 3.86, SD = 1.38), t (13) = -2.93, p = .01, d = .96. In addition, teachers’ ratings for their collaboration with colleagues showed significant progress at posttest ( M = 4.21, SD = 1.28), compared to pretest ( M = 3.18, SD = 1.07), t (13) = -2.99, p =.01, d = .87. We observed teacher behavior ( n =13, teacher 3 was absent for maternity leave) during their enactment at pretest and posttest of the AHL steps on two parts: Implement intervention activities [Step 4] and Stimulate oral language [Steps 5 to 7]. Additionally, we observed the number of parents present at these parent-child activities at school. T-tests revealed significant improvements of teacher behavior on each of the four steps of the AHL. The effect sizes of d = 1.04 to 1.62 are large to very large effects (Cohen, 1994). Table 4.3 presents the results.