Research Question 1

Background:

Most of the past work on testing accommodations has looked at which students *get* accommodations but not which students *use* them. In a prior, already-published paper, our team found that only about 28% of those students with disabilities who received extended time on the 2017 NAEP 8th Grade Mathematics assessment actually used any of their additional time. We further found that of those students who received extended time accommodations, students who felt that the test was not important were less likely to use any additional time, and students who used text-to-speech functioning on the test were more likely to use additional time.

In this new work, we extended our past research in two ways. First, we examined additional predictors of use of extended time among eligible 8th graders. Second, we extended our project to 4th graders. Younger students could (in theory) be less strategic or knowledgeable about using accommodations and accessibility features (such as text-to-speech), but they might also be more motivated than older students to do well on a low-stakes test.

Methods:

We started our work on this question by examining predictors of 4th graders' eligibility for extended time accommodations on the 2017 NAEP mathematics test. We then examined predictors of 4th graders' use of extended time. The predictors that we used included:

(a) gender

(b) race/ethnicity

(c) free/reduced price lunch status (a proxy for socioeconomic status)

(d) disability category

(e) English learner status

(f) Use of text-to-speech (asking the computer to read text to them)

(g) Score on the test

The analyses predicting use of extended time had three additional predictors:

(h) Self-reported effort put towards the test

(i) Perception of how important the test was

(j) Self-reported motivation to do well in math

Because some students were from the same school, we also included school as a higher-level predictor in the eligibility analyses.

After analyzing the 4th grade data in the ways just described, we reanalyzed the 8th grade data using a very similar strategy. Finally, we combined the Grade 4 data with the Grade 8 data, and ran an analysis predicting use of extended time in the whole combined sample. This allowed us to determine if grade level predicted use of extended time when other variables were controlled for.

Results

Our findings included the following:

• In the Grade 4 and (especially) the Grade 8 data, which school a student attended played a substantial role in determining eligibility for extended time accommodations. 

• When the Grade 4 and Grade 8 data were combined, we were able to determine that fourth graders who were given extended time were more likely to use it, whereas eighth graders who were given extended time were less likely to use it.

• In the Grade 4 sample:

  • Students with certain disability classifications were more or less likely than peers with other disability classifications to be found eligible for extended time, even controlling for other variables:

    • Autistic students were more likely to be eligible

    • Students with specific learning disabilities were more likely to be eligible

    • Students with speech/language impairments were less likely to be eligible

  • In addition, English Learners were more likely to be eligible for extended time.

  • Among students found eligible for extended time, those who were either of Hispanic origin or female were more likely to use it. Students who used text-to-speech were also more likely to use their extended time. 

  • Finally, extended time use was positively associated with students' scores on the item block.

Implications

• Which school a student attends appears to have a substantial effect on whether the student receives extended time accommodations, even after accounting for variability in student features. This suggests that schools have different thresholds or processes for making decisions about extended time. A standardized decision-making process may be helpful for reducing this variability.

• Students classified as having speech/language impairments may have less need for extended time accommodations on average, but teams should ensure that they are making accommodation decisions based on each student's individual needs and not relying on the classification itself to determine the decision.

• On a low-stakes test (as NAEP is), fourth graders who are given extended time are more likely to use it than eighth graders are. This may be due to higher levels of test-taking motivation in fourth graders, and consistent with that explanation, we also found that fourth graders reported higher levels of effort on the test than eighth graders did. Eighth graders may need additional motivational incentives to perform their best on low-stakes tests.