Stiggins, R. J., Schmeiser, C. B., & Ferguson, R. L. (1978). Validity of the ACT assessment as an indicator of reading ability. Applied Psychological Measurement, 2(3), 339-346. https://journals.sagepub.com/doi/10.1177/014662167800200306

Validity of the ACT assessment as an indicator of reading ability

- Obtained demographics and scores of 9,344 students who took the ACT from 1972-75 at 9 local institutions from MSU 
- Ran multiple regression models for different combinations of test sections (all 4 sections – English, Math, Natural Science Reading, and Social Science Reading; no math section; only reading) 
- Compared ACT score combinations to Michigan State University Reading Test 
- Calculated internal consistency reliability estimates for section/score combinations 

Validity of the ACT assessment as an indicator of reading ability - method

- Social studies and natural science reading passages produced the best prediction, with a median of 0.70 
- Adding in English scores helped the prediction, but adding in math section did not 
- Compared to MSU Reading Test, students’ ACT scores from the regressions were similar
- Reliability of the natural science and social studies reading passages were was high 0.92 (0.95 if English section added)

Validity of the ACT assessment as an indicator of reading ability - results

- Although all the combinations considered produced reliable predictions of reading comprehension performance, ACT composite scores predicted better than scores of sub-sections or any other combination of sections
- Correlation of 0.70 for natural science and social studies reading passages indicates that outside of reading comprehension skills, other factors do affect student performance on the ACT

Validity of the ACT assessment as an indicator of reading ability - discussion

This study examined whether combinations of ACT section scores could predict reading performance well enough to inform placement in intensive reading programs. It asked three questions: Which ACT score combinations predict reading performance most efficiently? Does prediction accuracy vary among examinee subgroups? How reliable are the different score combinations?

University of Michigan - Ann Arbor

Google

This branch reviews empirical studies on the American College Testing (ACT) Program's validity as a measure of reading comprehension and its use for student placement and outcome prediction. Two studies are detailed in child nodes: an estimation of reading grade-level equivalents from the ACT English and Social Studies Reading sections, and an analysis of the ACT's validity as an indicator of reading ability. Additional studies referenced in this literature (not yet expanded into child nodes) examine the ACT's validity for developmental students over a four-year college program, reading and language abilities of college freshmen, predictive validity for students from low socioeconomic backgrounds, and comparisons of the ACT with the Descriptive Tests of Language Skills for freshmen in basic-skills programs.

Literature on ACT Reading Comprehension

Researchers wanted to see if the ACT English and Social Studies Reading sections are reliable predictors of performance on the Nelson-Denny Reading Test Forms C and E. They found that the ACT is a good indicator of reading ability and can be used for placement purposes. 

Learn Before

Related

Learn After