With regard to the heterogeneity of students learning in inclusive classes, there has been a growing focus upon teaching practices. In that line, it is recommended to adapt the instruction to students’ individual needs. This approach is consistent with the idea that the effectiveness of instruction may depend on the fit between instruction and students’ abilities, interests, and learning profiles (e.g., Meijer, 2001). Differentiated instruction means that teachers adapt their instructional style, learning materials, and learning time according to the students’ characteristics.

To test the effectiveness of differentiated arrangements, it is required to measure the extent of differentiated instruction. Common methodological approaches to assess teaching practices consider three potential perspectives: students, teachers, and external observers. Fauth, Decristan, Rieser, Klieme, and Büttner (2014) found, that students, teachers, and external observers diverge in their ratings of the teaching in third grade classrooms. These findings raise the question which data sources are most adequate to provide information about differentiated instruction. On the one hand, the ratings of external trained observers are seen as more objective than students or teachers who are directly involved. On the other hand, ratings of teachers and students may be more reliable because their ratings refer to a broader range of learning situations.

Concerning students’ ratings, another aspect may be relevant: Typically, low performing students or students with special educational needs get in touch with forms of differentiated instruction more often than students with average or high abilities (Henke, Bosse, Lambrecht, Jäntsch, Jaeuthe & Spörer, 2017). In this respect, rating differences between these groups of students can be expected.

The present longitudinal study examined teaching practices in inclusive classes and compared the ratings of primary school students and external observers toward differentiated instruction. The goal was to answer the following research questions: (1) To what extent do the ratings of differentiated instruction coincide between students and external observers? (2) Do ratings of students and external observers vary over time? (3) Is there a difference between students’ ratings depending on their level of academic achievement?

The data came from a German longitudinal study. The first measurement point was in the second term of fourth grade (t1).The next (t2) was in the first term of grade 5, and the last one (t3) in the second term of grade 5. The interval between two measurement points was approximately six months. The sample consisted of N = 233 students from ten inclusive primary schools. At each measurement point, students’ ratings of differentiated instruction were assessed with a self-report questionnaire. Here, students were asked to rate the teaching practices that their math and German teachers use (e.g. “In math, we can choose between different tasks”) on a four-point Likert scale ranging from 0 (strongly disagree) to 3 (strongly agree). The reliability of the scale was satisfying, both for German (Cronbach’s αt1 = .71, αt2 = .76, αt3 = .75) and math (Cronbach’s αt1 = .61, αt2 = .60, αt3 = .66). The students also completed standardized tests in math (Haffner, Baro, Parzer, & Resch, 2005) and German (Lenhard & Schneider, 2006). Prior to the questionnaires and tests, external observers attended ten German and ten math lessons per class and measurement point. In preparation of the observations, all observers were trained on video footage showing real life classroom situations. They learned to relate observation situations to defined categories and to handle software that was developed to document the observations. Each observation of a 45-minutes-lesson was divided into two parts: After 20 minutes and after 40 minutes, the observers had to assess the quantity of differentiated instructions, measured by five items (Helmke, 2010) on a four-point Likert scale, e.g. “The children can choose between different tasks”. The reliability of the scale was satisfying, both for German (Cronbach’s αt1 = .90, αt2 = .74, αt3 = .82) and for math (Cronbach’s αt1 = .86, αt2 = .75, αt3 = .84). Regarding the ratings of the external observers, for each class a mean score for differentiated instruction was calculated for both subjects, and for each measurement point. In a second step, those scores were summarized per measurement point as an aggregated mean score for differentiated instruction that had been observed within a class. The procedure was the same for math and German. To assess how similar observers’ and students’ ratings were, the final observation parameters were correlated with the students’ assessments at the classroom level. Intra-class correlations were computed.

Three main results turned out: (1) Compared to the students’ rating of differentiated instruction, the observers’ ratings were significantly lower. This result was found for all three measurement points and both subjects, German and math. The ratings of the external observers ranged from 0.68 to 0.82, and were thus below the theoretical scale mean of 1.50. The students’ ratings ranged from 1.61 to 1.83. Furthermore, the analyses revealed high values for the standard deviations and for the intra-class correlations indicating substantial differences between classes with respect to differentiated instruction. (2) Whereas the ratings of the external observers were stable over time, students’ ratings of differentiated instruction varied between measurement points. The intra-class correlations diminished over time, from 0.42 to 0.19. The ratings of the students are measurements on an interpersonal level, understood as assessments of different persons. Thus, the interpretation of the intra-class correlation is that students perceived differentiated instruction varying from an individual point of view. (3) For German, the correlations of students’ academic achievement and ratings of differentiated instruction showed a specific pattern of results. Academic achievement at t1 (t2, respectively) was negatively correlated with the rating of differentiated instruction at t2 (t3, respectively). The lower the academic achievement of a student at an earlier measurement point was, the more differentiated instruction he or she perceived at a later measurement point. For math, academic achievement measured at t2 and differentiated instruction measured at t3 were negatively correlated. Same construct – different perspectives? The results will be discussed in light of adequate assessment of teaching quality, especially in heterogeneous classes (Decristan, Fauth, Kunter, Büttner, & Klieme, 2017; Praetorius, Lenske, & Helmke, 2012).

Decristan, J., Fauth, B., Kunter, M., Büttner, G., Klieme, E. (2017). The interplay between class heterogeneity and teaching quality in primary school. International Journal of Educational Research, 86, 109–121. DOI: https://doi.org/10.1016/j.ijer.2017.09.004. Fauth, B., Decristan, J., Rieser, S., Klieme, E., Büttner, G. (2014). Student ratings of teaching quality in primary school: Dimensions and prediction of student outcomes. Learning and Instruction, 29, 1–9. DOI: https://doi.org/10.1016/j.learninstruc.2013.07.001. Haffner, J., Baro, K., Parzer, P., & Resch, F. (2005). HRT 1-4: Heidelberger Rechentest [Heidelberg numeracy test]. Göttingen: Hogrefe. Helmke, A. (2010). Unterrichtsqualität und Lehrerprofessionalität [lesson quality and teaching professionality] (3. Auflage). Seelze-Velber: Klett/Kallmeyer. Henke, T., Bosse, S., Lambrecht, J., Jäntsch, C., Jaeuthe, J., Spörer, N. (2017). Mittendrin oder nur dabei? Zum Zusammenhang zwischen sonderpädagogischem Förderbedarf und sozialer Partizipation von Grundschülerinnen und Grundschülern [Primary school children with special educational needs and their social participation]. Zeitschrift für Pädagogische Psychologie [German Journal of Educational Psychology], 31(2), 111–123. DOI: https://doi.org/10.1024/1010-0652/a000196. Lenhard, W. & Schneider, W. (2006). ELFE 1-6: Ein Leseverständnistest für Erst- bis Sechstklässler [reading comprehension test for grade 1-6]. Göttingen: Hogrefe. Meijer, C. J. W. (2001). Inclusive Education and Effective Classroom Practices. European Agency for Development in Special Needs Education, Odense. Praetorius, A.-K., Lenske, G., Helmke, A. (2012). Observer ratings of instructional quality: Do they fulfill what they promise? Learning and Instruction, 22(6), 387–400. DOI: https://doi.org/10.1016/j.learninstruc.2012.03.002.