(Post last updated June 14, 2022)
Review panel summary
The original Math-Up Skills Test (MUST) is a 16 question, free-response math instrument that was developed to assess students’ ability to conduct basic pre-college mathematical operations including the following topics: multiplication, division, fractions, scientific and exponential notation, logarithms, square roots, and balancing chemical equations [1]. An updated version of MUST was subsequently developed after the pilot study was conducted and was expanded to contain 20 questions, which includes the following additional topics: simplification of a complex fraction, division by zero, simplification of a mixed operations fraction, and calculation of fraction decimal equivalent [2]. In Texas where the MUST was developed, some of these math topics are introduced as early as fourth grade and the remaining topics are covered by eleventh grade [1]. Students take 15-minutes to complete the MUST in-person without the use of a calculator. The MUST is graded by hand, and for each correct answer, a score of 1 point is awarded and no points are awarded for an incorrect answer. An updated version of the MUST instrument and its answer key is readily available in the supplementary information [2].
There are two versions of the MUST. In terms of single administration reliability, both versions of MUST show good evidence of reliability. Test content validity evidence was provided by professors from multiple disciplines (mathematics, chemistry, chemistry education) and an agreement of 100% was reached on correct answers to the items [2]. Validity evidence of relation to other variables was repeatedly established with correlation and regression analyses [1 - 6].
The MUST has been evaluated with students enrolled in general chemistry [1 - 6], and engineering courses [1] across various different types of institutions (public and private universities and Hispanic Serving Institutions (HSI)), specifically in Texas, USA.
Recommendations for use
The MUST is developed to assess students’ ability to perform simple pre-college math problems. According to validity aspects reported in literature, the data collected with this instrument have shown good evidence of relations to other variables. According to the reliability aspects reported in literature, single administration reliability based on KR-20, 21 and coefficient alpha values are consistently high throughout the analyzed studies [1 - 6]. Users should be aware that the instrument is open response, and must be graded by hand. This might cause errors introduced by graders or be a time burden on the graders.
Since the instrument has good predictability for success in general chemistry, instructors might consider using MUST scores (in conjunction with other information) for advising purposes. However, no evidence for the discrimination and difficulty of items on the instrument are provided. In the studies reviewed by the panel [1 - 6], the MUST has been used only for observational or informational purposes, and not to make advising/placement decisions or restrict student access to coursework, therefore the collection of consequence testing validity evidence is recommended before using the MUST in high stakes decisions. Users should be aware that the MUST was developed based on Texas’ K-12 math and chemistry standards and has only been used in Texas, albeit at a variety of institutions.
Details from panel review
The developers and users of MUST reported some aspects of validity (relation to other variables). Particularly, medium to strong correlations (r = 0.288 - 0.542) were reported between MUST score and course grades in general chemistry [1 - 6] and engineering chemistry [1] and MUST score and quantitative literacy and quantitative reasoning [5] (r =0.60). In addition, a relation was found between MUST score and Test of Logical Thinking (TOLT) scores where higher scores on both coocurred as did lower scores on both [4]. Reliability evidence (single administration reliability) for both versions of the instrument based on KR-20, 21 and coefficient alpha are consistent throughout the papers analyzed (KR-21 = 0.821 to 0.855; coefficient alpha = 0.85) [1 - 6].
Although there is no formal information provided for the difficulty of items on MUST, there is sufficient information provided to calculate the difficulty based on given mean scores for each topic assessed on the MUST [1]. Regression analyses have demonstrated the ability for MUST scores to be predictive of success in general chemistry [1-4], specifically when coupled with other variables (i.e., quantitative reasoning, quantitative literacy) [5] and other demographic factors [6]. When the MUST instrument was coupled with quantitative reasoning and quantitative literacy instruments, up to 50% of chem I students and about 45% of chem II students who will not succeed has been predicted [5]. It is also interesting to note that when the MUST instrument has been used with ACS exam and demographic variables, it provided an even more accurate predictor of success (83.4%) [6].
Lastly, MUST has been used at multiple institutions (R1, private, public, HSI) and it is encouraging to see that similar results were gathered across all studies. However, all of the studies were conducted in Texas [1 - 6]. Therefore, the results cannot necessarily be generalized at the national level. While the similar results are encouraging, the addition of evidence supporting response process validity would further strengthen the interpretation of MUST data.
References
[1] Albaladejo, J.D.P., Broadway, S., Mamiya, B., Petros, A., Powell, C.B., Shelton, G.R., Walker, D.R., Weber, R., Williamson, V.M., & Mason, D. (2018). ConfChem Conference on Mathematics in Undergraduate Chemistry Instruction: MUST-Know Pilot Study–Math Preparation Study from Texas. Journal of Chemical Education, 95(8), 1428-1429. https://doi.org/10.1021/acs.jchemed.8b00096
[2] Williamson, V.M., Walker, D.R., Chuu, E.,Broadway, S., Mamiya, B., Powell, C.B., Shelton, G.R., Weber, R., Dabney, A.R., & Mason, D. (2020). Impact of basic arithmetic skills on success in first-semester general chemistry. Chemistry Education Research and Practice, 21(1), 51-61. https://doi.org/10.1039/C9RP00077A
[3] Powell, C.B., Simpson, J., Williamson, V.M., Dubrovskiy, A., Walker, D.R., Jang, B., Shelton, G.R., & Mason, D. (2020). Impact of arithmetic automaticity on students’ success in second-semester general chemistry. Chemistry Education Research and Practice, 21(4), 1028-1041. https://doi.org/10.1039/D0RP00006J
[4] Alivio, T.E.G., Howard, E., Mamiya, B., & Williamson, V.M. (2020). How does a math review impact a student’s arithmetic skills and performance in first-semester general chemistry? Journal of Science Education and Technology, 29, 703-712. https://doi.org/10.1007/s10956-020-09851-7
[5] Sheldon, G.R., Mamiya, B., Weber, R., Walker, D.R., Powell, C.B., Jang, B., Dubrovskiy, A.V., Villalta-Cerdas, A., & Mason, D. (2021). Early warning signals from automaticity diagnostic instruments for first- and second-semester general chemistry. Journal of Chemical Education, 98, 3061-3072. https://doi.org/10.1021/acs.jchemed.1c00714
[6] Willis, W.K., Williamson, V.M., Chuu, E., & Dabney, A.R. (2022). The relationship between a student’s success in first-semester general chemistry and their mathematics fluency, profile, and performance on common questions. Journal of Science Education and Technology, 31, 1-15. https://doi.org/10.1007/s10956-021-09927-y