Measures of long-term trends in mathematics: linking large-scale assessments over 50 years

Majoros, Erika; Rosén, Monica; Johansson, Stefan; Gustafsson, Jan‐Eric

doi:10.1007/s11092-021-09353-z

Cited by 7 publications

(2 citation statements)

References 30 publications

(45 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Researchers have been keen on equating test scores to large-scale data assessment in the past decade. Majoros et al (2021) and Strietholt & Rosén (2016) use data from the IEA (International Association for the Evaluation of Educational) for mathematical literacy data. Similarly, equating test scores for the TIMSS and PIRLS assessment programs on reading and mathematics literacy between different countries (Khorramdel et al, 2022).…”

Section: Introductionmentioning

confidence: 99%

The Use of Stocking-Lord and Haebara Methods in Horizontal Equating: A Case of Indonesian Madrasah Competence Assessment

Kusaeri,

Ridho,

Wahyudi

2024

JP3I

View full text Add to dashboard Cite

Indonesian Madrasah Competence Assessment (AKMI) is a national assessment implemented each year held by the Ministry of Religious Affairs. One of the uniqueness of the AKMI is the use of different tests every year. AKMI focuses on capturing the development of learning in Madrasa by comparing the test scores of the current year with the previous year. An equating process is crucial for valid results when comparing scores. Therefore, this research aims to (a) equate the scientific literacy assessment tools at AKMI in 2022 with 2023 and (b) evaluate the business process of developing AKMI scientific literacy instruments (along with the MSAT design), which has implications for the equating process. This study adopted a Non-Equivalent Anchor Test (NEAT) design because the two test sets were parallel years, and the participants were from a diverse population. The data is from the AKMI Science Literacy of the Ministry of Religious Affairs, with 303,987 participants in 2022 and 342,987 in 2023 from the Islamic elementary school level. A total of 674 scientific literacy instrument items in 2022 and 1,392 items in 2023, with 90 items used as anchor items. There are 3 stages of analysis: pre-equalization, equalization calibration, and post-equalization analysis. The results show that there are differences in item parameter estimation results between 2022 and 2023, where 2022 has a higher level of item difficulty. Furthermore, the Stocking-Lord and Haebara methods had proven to be effective and had produced estimates with minimal differences in the equating process. In addition, the anchor items used as the basis for the equating do not represent the items as a whole in the item pool. These findings indicate the need for firm, careful standardization based on psychometric principles of the process at AKMI, from developing items to assembling items, testing, determining anchor items, and assembling items in the MSAT application.

show abstract

Section: Introductionmentioning

confidence: 99%

The Use of Stocking-Lord and Haebara Methods in Horizontal Equating: A Case of Indonesian Madrasah Competence Assessment

Kusaeri,

Ridho,

Wahyudi

2024

JP3I

View full text Add to dashboard Cite

show abstract

“…They also enter implicitly in our sample selection criteria described in Section 4 (below). A recent approach to more reliable linking of the different international tests is found inMajoros, Rosén, Johansson, and Gustafsson (2021).7 For a discussion of the measurement issues, seeHanushek et al (2022).…”

mentioning

confidence: 99%

Long-run Trends in the U.S. SES—Achievement Gap

Hanushek

Light

Peterson

et al. 2022

Education Finance and Policy

View full text Add to dashboard Cite

Rising inequality in the United States has raised concerns about potentially widening gaps in educational achievement by socioeconomic status (SES). Using assessments from LTT-NAEP, Main-NAEP, TIMSS, and PISA that are psychometrically linked over time, we trace trends in SES gaps in achievement for U.S. student cohorts born between 1961 and 2001. Gaps in math, reading, and science achievement between the top and bottom quartiles of the SES distribution have closed by 0.05 standard deviation per decade over this period. The findings are consistent across alternative measures of SES and subsets of available tests and hold in more recent periods. At the current pace of closure, the achievement gap would not be eliminated until the second half of the 22nd century.

show abstract

Comparison of Studies: Comparing Design and Constructs, Aligning Measures, Integrating Data, Cross-validating Findings

Klieme

2022

International Handbook of Comparative Large-Scale Studies in Education

View full text Add to dashboard Cite

Measures of long-term trends in mathematics: linking large-scale assessments over 50 years

Cited by 7 publications

References 30 publications

The Use of Stocking-Lord and Haebara Methods in Horizontal Equating: A Case of Indonesian Madrasah Competence Assessment

The Use of Stocking-Lord and Haebara Methods in Horizontal Equating: A Case of Indonesian Madrasah Competence Assessment

Long-run Trends in the U.S. SES—Achievement Gap

Comparison of Studies: Comparing Design and Constructs, Aligning Measures, Integrating Data, Cross-validating Findings

Contact Info

Product

Resources

About