Jens Möller

The classic internal/external frame of reference (I/E) model describes the effects of students’ math and verbal achievements on their math and verbal self-concepts. It assumes that social (external) comparisons (with achievements of others) are reflected in positive within-domain effects, whereas dimensional (internal) comparisons (with achievements in other domains) are reflected in negative between-domain effects. In recent decades, these assumptions have received broad empirical support. However, Wolff (2021) argued that the positive within-domain effects in the classic I/E model may also reflect dimensional comparison effects, leading to an overestimation of the impact of within-domain achievements (including social comparisons). To illustrate this argument, he proposed the difference I/E model as a reparameterization of the classic I/E model, in which self-concepts in a given domain are regressed on the difference between achievements in the same and the other domain, rather than on absolute achievement in the other domain, to draw inferences about the strength of dimensional comparison effects. To date, however, a systematic comparison of the two I/E model specifications is still lacking. In this research, we therefore compared the two I/E model specifications across six empirical studies. Overall, these studies provided evidence consistent with the assumption that the within-domain effects in the classic I/E model (but not in the difference I/E model) also reflect between-domain achievement effects (including dimensional comparison effects) and that relative achievement information is salient when students evaluate their self-concepts, thereby supporting the theoretical relevance of the difference-score parameterization. Implications for prior and future I/E model research are discussed.

Assessing texts from pupils learning English as a foreign language at elementary school poses various challenges, one being the distorting influence of orthography, vocabulary and punctuation on the assessment of other criteria. With a pre-post control group design, this study investigated whether the existence of these halo effects can be confirmed, and whether they can be reduced through repeated halo prompts. 284 pre-service teachers assessed six texts each, three in the pre-test and three in the post-test. The texts were randomly drawn from a larger text sample from a sixth-grade corpus. Participants were randomly assigned to one of three experimental groups (receiving halo prompts cautioning against orthography-, vocabulary- or punctuation-induced halo effects) or to the control group (neutral prompt). Multi-level regression analyses showed that orthography- and vocabulary-induced halo effects exist and that they could be significantly reduced by the interventions. Including halo prompts in the assessment process can therefore help to improve the quality of the assessment. The occurrence of punctuation-induced halo effects could not be confirmed. Teachers, teacher educators and language testing professionals should be aware that orthography and vocabulary can distort EFL writing assessment at elementary school level, and that awareness of this bias can reduce its impact.

Implicit academic self-concepts and intrinsic values describe students’ automatically activated perceived abilities and attitudes towards school subjects. In the present research, we examined these implicit constructs within the context of a framework that integrates key elements of three prominent theories from the field of educational psychology: expectancy-value theory, dimensional comparison theory, and theory on vocational interests. To this end, we developed four go/no-go association tasks (GNATs) to measure implicit math and German self-concepts and values. Furthermore, we examined the formation and impact of these implicit constructs using cross-sectional data from a sample of 946 ninth- and tenth-graders from German high schools. Guided by three central research questions, we found (1) evidence for the validity of the GNATs to measure students’ implicit self-concepts and values in math, (2) that implicit math self-concepts and values appear to be influenced by dimensional comparisons (where students compare their achievements in different subjects), but not by social comparisons (where students compare their achievements with those of other students), and (3) that implicit math self-concepts predict students’ vocational interests over and above their achievements, explicit self-concepts, and explicit values in math and German. Our research also demonstrated the benefit of combining key elements from different theories of educational psychology to gain a deeper understanding of the relationships between students’ achievements, self-concepts, values, and vocational interests. Moreover, it is of great practical relevance as it highlights the importance of implicit self-concepts and values for students.

By quantitatively integrating results of meta-analyses, second-order meta-analyses (SOMAs) from educational psychology address far-reaching questions of interest to adjacent scientific fields, policy, and educational practice. Conducting a SOMA, however, is time-consuming, particularly the extraction of statistical results from meta-analyses. Large Language Models (LLMs) have the potential to accelerate data extraction, as evidenced by evaluations of data extraction from primary studies. Evaluations of their accuracy in extracting results from meta-analyses to conduct SOMA are needed. We compare the accuracy of three LLMs with that of human experts extracting data from 156 educational meta-analyses investigating students’ achievement, randomly chosen from the Visible Learning database. LLMs and humans extracted effect sizes, the number of included studies, and the number of students. To assess accuracy, we established a gold standard benchmark by resolving discrepancies between LLMs and human experts through expert adjudication. Results show that the distributions of data from all coders and the standard were similar and showed no systematic bias. Accuracy reached ICCs of 0.95/0.81 for the two human experts, and 0.96/0.97/0.96 for LLMs (Gemini2.5Pro/GPT4.1/GPTo3), with percentage agreement of 86%/80% (humans) and 81%/78%/77% (LLMs). The results demonstrate that LLMs achieve data extraction accuracy comparable to that of human experts. We discuss the conditions under which data extracted by LLMs, humans, or a hybrid of both can be considered validated for use in educational SOMAs.

The internal/external frame of reference (I/E) model posits that students’ academic self-concepts (ASCs) in math and verbal subjects are shaped by social comparisons (interindividual with others) and dimensional comparisons (intraindividual across subjects). Extending prior research, this study examines (1) whether the I/E model assumptions apply to psychometrically measured and self-estimated numerical and verbal intelligence and (2) whether measured intelligence and math and German grades uniquely predict self-estimated intelligence and math and German ASCs within and across domains. Data from 456 German secondary school students were analyzed using structural equation modeling. Results showed positive social and negative dimensional comparison effects of numerical and verbal intelligence on self-estimated numerical and verbal intelligence, which largely persisted after controlling for grades. Furthermore, math and German grades demonstrated unique social and dimensional comparison effects on self-estimated intelligence beyond measured intelligence. Similarly, measured intelligence and grades uniquely predicted math and German ASCs in line with the I/E model assumptions. The findings (1) support the I/E model’s applicability to measured and self-estimated numerical and verbal intelligence, thus informing the formation of intelligence self-estimates; (2) highlight the unique contributions of intelligence and grades to shaping self-estimated intelligence and ASCs; and (3) shed light on how dimensional comparison effects on self-estimated numerical and verbal intelligence contribute to and reinforce students’ beliefs that one is either a “math person” or a “verbal person”.

Teachers' assessments of students' writing are key for adapting instruction to individual needs but are often influenced by biases. The present studies examined whether student characteristics – gender and migration background – affect text assessments in the digital tool Student Inventory. In Study 1 (n = 117), pre-service teachers assessed six texts differing in quality and students' gender. In Study 2 (n = 127), students' migration background was manipulated instead. Texts were assessed holistically and analytically, considering content, style, and linguistic accuracy. Results from both studies showed that participants reliably differentiated between low- and high-quality texts on all assessment scales. The studies revealed small significant gender biases only for medium-quality texts. The findings suggest that the Student Inventory facilitates an objective assessment of students' writing and supports fair assessment practices, regardless of students' gender or migration background.

An expanding literature has explored the impact of second language (L2) self-concept on language achievement and motivation. This review adopts a psychological lens to examine L2 self-concept, beginning with an outline of prominent theoretical frameworks: situated expectancy‐value theory; Shavelson et al.’s (1976) hierarchical, multidimensional model of self-concept; and the integrated Marsh/Shavelson model (Marsh et al., 1988). Building on these foundations, we introduce Marsh’s (1986) internal/external (I/E) frame of reference model, which posits that, within domains such as mathematics or verbal subjects, achievement and academic self-concept are positively correlated, whereas achievement in one domain relates negatively to self-concept in a different domain. Originally confined to mathematics and a single verbal subject in the firstlanguage (L1), the I/E model has been extended to additional disciplines, including L2, through dimensional comparison theory (DCT; Möller & Marsh, 2013). Central to this paper is a synthesis of findings from a recent meta-analysis investigating I/E effects that incorporate measures of L2 achievement and self-concept. Results demonstrate that higher L2 achievement bolsters L2 self-concept while simultaneously diminishing self-concepts in math, physics, biology, and L1; conversely, accomplishments in those other fields exert negative influences on L2 self-concept. We conclude by discussing both theoretical insights and practical applications of self-concept research for language education generally, and L2 instruction specifically.