Changes in Effects of Formal Teacher Competence on 4th graders' Reading Achievement between 2001 and 2011 in Sweden

Author(s):

Eva Myrberg(presenting / submitting)Stefan Johansson Monica Rosén

Conference:

ECER 2015

Network:

09. Assessment, Evaluation, Testing and Measurement

Format:

Paper

Session Information

09 SES 03 A, Findings from International and National Large-Scale Assessments: Relating Teacher Variables to Student Achievement

Paper Session

Time:

2015-09-08

17:15-18:45

Room:

326. [Main]

Chair:

Inga Arffman

Contribution

Background

Student social background was long considered the one prominent factor affecting students’ school achievement. Only since the late nineties, research on effects of teacher competence on student achievement has started to accumulate. There is an emerging consensus that teachers vary substantially in their ability to enhance student learning (Hattie, 2009). However, the understanding of if and how teacher education affect the development of professional knowledge is still limited (Cochran-Smith & Zeichner, 2005) some maintaining that teacher competence is a personal trait that cannot be measured by observable variables (Hanushek, , 2003). In recent years, though, in the US, as well as in Europe, successful attempts have been made to relate teacher education to student achievement (Baumert et al., 2010; Darling-Hammond, 2006; Nye, Konstantopolous & Hedges, 2004).

There are large difference between countries with regard to length, structure and contents of teacher education. Obviously, this calls for country-specific analyses based on accurate information of the specific features of particular teacher education programs. As proposed by Croninger, King Rice, Rathbun and Nishio (2007), more refined measures of teachers’ preparation tend to be better predictors of student achievement than more conventional and broader measures like certification status.

For the part of Sweden, Myrberg (2007) studied effects of teacher competence and estimated substantial effects of education directed towards the specific grade and the specific subject on 3^rd graders’ reading achievement levels. Frank (2009) studied the relationship between teachers with different levels of formal competence and student achievement in 3^rd grade. She estimated strong positive effects of formal competence with respect to preparation for subject and grade on student reading achievement. Johansson, Myrberg & Rosén, (2014) further developed Franks’ (2009) model of formal competence and estimated considerable effects on students’ reading achievement, measured both by reading test scores and teachers’ assessment of student reading achievement levels.

In Sweden, teacher education has been revised several times during the past five decades and content knowledge as well as pedagogical content knowledge has become less pronounced (Högskoleverket, 2005). From 1988 and with a peak in 2001, the profession has developed into one single, flexible teacher category with less focus on specialized subject knowledge (Stenlås, 2009; SOU, 2014). Changes in teacher education curriculum are likely to affect teacher effectiveness. Furthermore teacher effects on student achievement probably interacts with students background factors (Nye, Konstantopolous & Hedges, 2004). In international literature the notion of peer effects has strong support and classroom composition with respect to student social background has proven to be a strong predictor of school success (Hattie, 2009; Hoxby, 2000). The influence of socio-economic background on student achievement has increased markedly in Sweden during latter decades (Yang-Hansen & Gustafsson, 2013).

In summary, according to previous research, students’ social background and teacher competence are two important predictors of student achievement. Changes in their respective relative weight of influence on individual and group levels are of great interest. So is the possibility of interaction between these two variables.

Aim

The main aim is to investigate changes in effects of teacher competence on student reading achievement. Four questions have been formulated: Can effects of teacher competence on student reading achievement be estimated for 4th grade students in Sweden? Are there differences in effects of teacher competence for 4^th graders over time? Are there changes in SES- effects on student reading achievement over time on individual and group level? Do student SES and teacher competence co-vary and are there any changes over time?

Method

Data and method The empirical base consists of data from IEA’s regularly recurring reading study, PIRLS 2001, 2006, and 2011 (Progress in International Reading Literacy Study). Well over 30 countries have participated in the PIRLS studies, which assesses the students’ reading ability in grade 4. A national extension in Sweden provides rich data on teachers’ education. Given the aforementioned reforms of teacher education in Sweden, teachers are holding different certificates and have experienced varying amount of focus on, for example, reading pedagogy and Swedish language during education, which is also reflected in data. Four variables: type of teacher education, amount of reading pedagogy and Swedish language together with years of teacher experience define the latent variable “formal teacher competence” in our study. Besides formal teacher competence, information about students’ reading achievement, student socio-economic status (SES) and time point for PIRLS assessment are used in the analyses. The information regarding SES is collected from the parent questionnaire. The main method of analysis is multilevel Structural Equation Modeling (SEM) with latent variables. These analytical techniques have developed dramatically in the past few years. Educational assessment data often has a nested observational structure, students being clustered in classrooms, classrooms being clustered in schools and so on. Such hierarchically structured data is difficult to analyze, but starting in the 1980s, appropriate analytical methods have been developed through extensions of the basic regression model (e.g., Snijders & Bosker, 1999). The analyses are conducted using the computer program Mplus (Muthén & Muthén, 1998-2012). The analyses were first run for each time-point respectively. The SES variable was related to student achievement at both individual and class-level. Since only one teacher has been selected for each classroom, formal teacher competence could only be introduced at the class-level. The competence variable was used as an independent variable to investigate whether it was possible to find any influence on student achievement at class-level (students’ average achievement in one classroom). In addition, class-SES was used to investigate if teachers tend to, by self-selection, be concentrated to classrooms with high-SES students and/or high-SES students tend to attend classrooms with more adequately educated teachers. In the final analysis, a pooled dataset was used and a dummy-variable for year of study was used to investigate if the estimates differed significantly between the three time-points.

Expected Outcomes

Results First, a common measurement model for years 2001, 2006, and 2011 of 1) teacher competence and 2) students’ SES was fitted to the data. The models obtained acceptable to good fit. Thereafter, a two-level SEM was formulated where SES and student achievement constituted the relation at the individual level. At the class-level, SES and formal teacher competence was related to student achievement. SES and teacher competence were correlated in order to account for potential selection effects. In 2001, formal teacher competence had a significant relationship to student achievement. The effect was estimated to about .23, which must be considered as fairly high in this context. In many cases, teachers had not taught their students for more than about 6 months. No correlation between SES and FTC was found, indicating no selection effects. In 2006, FTC did not have any significant relation to student achievement. There was a significant correlation between SES and FTC (.28), indicating that teachers with higher FTC had classrooms with higher levels of SES. In 2011, the relation between FTC and achievement and FTC and SES was non-significant. At individual level, the effect of SES on students’ achievement increased from 0.25 in 2001, to .33 in 2006, and to .36 in 2011. At class-level the relationship was about .80 both 2001 and 2006, whilst in 2011, it increased to .87. The entering of two dummy variables representing year of study showed that the differences between 2001, 2006 and 2011 were significant. We conclude that the SES influence on student achievement has increased substantially in the past few years in Sweden, whereas the effects of teacher competence found in 2001 have disappeared. The changed social composition between schools between the years 2001 and 2011 along with changes in teacher competence have negatively affected students’ reading achievement levels.

References

References Baumert, J., Kunter, M., Blum, W., Brunner, M., Voss, T., Jordan, A., . . . Tsai, Y.-M. (2010). Teachers’ Mathematical Knowledge, Cognitive Activation in the Classroom, and Pupil Progress. American Educational Research Journal, 47, 133-180. Cochran-Smith, M & Zeichner, K. M. (2005). Teacher education, The report of the AERA Panel on Research and Teacher Education. Mahwah, NJ: Lawrence Erlbaum. Croninger, R. G., King Rice, J. K., Rathbun, A. & Nishio, M. (2007). Teacher qualifications and early learning: Effects of certification, degree, and experience on first-grade student achievement. Economics of Education Review, 26, 312-324. Darling-Hammond, L. (2006). Powerful Teacher Education. San Francisco: Jossey Bass. Frank, E. (2009). Läsförmågan bland 9-10-åringar. Betydelsen av skolklimat, hem- och skolsamverkan, lärarkompetens och elevers hembakgrund. ACTA: Gothenburg. Retrieved from http://hdl.handle.net/2077/20083. Hanushek, E. A. (2003). The Failure of Input-Based Schooling Policies. The Economic Journal, 113, 64-98. Hattie, J. A. (2009). Visible learning, A synthesis of over 800 meta-analyses relating to achievement. New York: Routledge. Hoxby, C. (2000) Peer effects in the classroom: Learning from gender and race variation. NBER Working Paper 7867. Högskoleverket (2005) Reformuppföljning och kvalitetsbedömning av den nya lärarutbildningen vid svenska universitet och högskolor. Stockholm: Högskoleverket. Johansson, S., Myrberg, E. & Rosén, M. (2014) Formal Teacher Competence and its Effect on Pupil Reading Achievement. Scandinavian Journal of Educational Research. Muthén, L. K., & Muthén, B. O. (1998-2012). Mplus User's Guide. Los Angeles, CA: Muthén & Muthén. Myrberg, E. (2007). The Effect of Formal Teacher Education on Reading Achievement of 3rd-Grade Pupils in Public and Independent Schools in Sweden. Educational Studies, 33, 145-162. Nye, B., Konstantopoulos, S., & Hedges, L. V. (2004). How Large Are Teacher Effects? Educational Evaluation and Policy Analysis, 26, 237-257. Snijders, T.A.B., & Bosker, R.J. (1999). Multilevel analysis: An introduction to basic and advanced multilevel modelling, London: SAGE Publications. SOU (2014:15). Staten får inte abdikera – om kommunaliseringen av den svenska skolan. [The state must not abdicate – about the municipalization of the Swedish school]. Stenlås, N. (2009). En kår i kläm – Läraryrket mellan professionella ideal och statliga reformideologier. Stockholm: Regeringskansliet, Finansdepartementet. Yang Hansen, K., & Gustafsson, J-E. (2013). Changes in Educational Inequality in Sweden and Norway: Evidence from IEA PIRLS studies between 2001 and 2011. Paper presented at the European Conference on Educational Research (ECER), Istanbul, Turkey, September, 2013.

Author Information

Eva Myrberg (presenting / submitting)

University of Gothenburg

Gothenburg

Stefan Johansson

University of Gothenburg, Sweden

Monica Rosén