Reading Attitudes and U.S. PIRLS Performance 2021
PIRLS 2021 grouped students by how much they reported liking reading. U.S. reading achievement and student shares across the three attitude categories are shown with PIRLS international averages for context.
U.S. Reading-Attitude Profile
Reading Achievement by Reading Attitude
Mean PIRLS reading scores for the three Students Like Reading categories
U.S. results refer to the fourth-grade cohort assessed at the beginning of fifth grade in fall 2021. Standard errors for the estimates appear in the data table.
IEA PIRLS 2021 International Results in Reading, Exhibit 7.1.
Full Reading-Attitude Data
Achievement values are PIRLS reading scale scores. Values in parentheses are standard errors.
| Very Much Like Reading | 25% (1.4) | 559 (8.6) | 42% (0.1) | 513 (0.5) |
| Somewhat Like Reading | 50% (1.4) | 557 (6.3) | 40% (0.1) | 501 (0.5) |
| Do Not Like Reading | 24% (1.4) | 530 (6.7) | 18% (0.1) | 491 (0.8) |
U.S. Achievement Was Nearly the Same Across the Two More Positive Attitude Groups
- U.S. students who very much liked reading averaged 559 points, while those who somewhat liked reading averaged 557. The two reported means were only 2 points apart. Exhibit 7.1 does not provide a pairwise statistical test for this U.S. comparison, so the values should be treated as a numeric difference rather than evidence that one group performed differently from the other.
- The U.S. Do Not Like Reading group averaged 530 points. That was 29 points below the Very Much Like Reading group and 27 points below the Somewhat Like Reading group on a numeric basis.
- The U.S. distribution was less concentrated in the most positive category than the PIRLS international average. Twenty-five percent of U.S. students were classified as Very Much Like Reading, compared with 42 percent internationally. NCES separately reported the U.S. 25 percent as lower than the international average.
- The international achievement averages declined numerically across all three categories, from 513 to 501 to 491. The U.S. pattern differed because the first two categories had nearly identical reported achievement before the lower score for the Do Not Like Reading group.
How PIRLS Defined Reading Attitudes
The Students Like Reading scale combines responses to eight statements about attitudes toward reading with two questions about reading activities outside school. The attitude items cover enjoyment, interest, discussion of reading, learning through reading, imaginative engagement, and related responses. One item about reading being boring is reverse coded.
The two activity questions measure how often students read for fun and how often they read to learn about topics that interest them. IEA combines the questionnaire responses using its context-scale methodology and assigns students to three categories.
| Scale Interpretation | ||
|---|---|---|
| Very Much Like Reading | 10.4 or above | Responses indicate a strongly positive pattern across the reading-attitude statements and frequent reading activity outside school. |
| Somewhat Like Reading | Above 8.3 and below 10.4 | Students whose scale scores fall between the two published cut points. |
| Do Not Like Reading | 8.3 or below | Responses indicate a less positive attitude pattern and less frequent reading activity outside school. |
The reported U.S. average on the Students Like Reading context scale was 9.4, with a standard error of 0.06. The scale was established in PIRLS 2016 with a centerpoint of 10; two scale points corresponded to one standard deviation in the distribution used to establish the scale.
Data Sources, Measurement and Comparability
Official Attitude and Achievement Data
The percentages, reading achievement scores, standard errors, attitude-scale average, and scale cut points come from IEA PIRLS 2021 International Results in Reading, Exhibit 7.1, Students Like Reading.
U.S. Reporting
NCES reports that 25 percent of U.S. students were classified as Very Much Like Reading, compared with an international average of 42 percent. The U.S. result was reported by NCES as lower than the international percentage.
Population and Timing
PIRLS normally assesses students near the end of fourth grade. Because U.S. data collection was delayed during the COVID-19 period, the United States assessed the fourth-grade cohort at the beginning of fifth grade in October–November 2021. Comparisons with countries assessed at the normal point in fourth grade therefore require attention to the timing difference.
U.S. Assessment Mode
The United States administered both digitalPIRLS and a paper-bridge assessment in 2021 and elected to report the paper-bridge achievement results. The bridge design linked the new digital administration with the established PIRLS scale.
Questionnaire Measurement
All assessed students were asked the PIRLS student questionnaire. The Students Like Reading scale is a multi-item measure rather than a response to a single question, so the three categories summarize a broader pattern of reading attitudes and out-of-school reading frequency.
Sampling Qualification
IEA flagged the United States as not satisfying its PIRLS 2021 sample-participation guidelines. Weighted U.S. school participation was 54 percent before replacement and 67 percent after replacement; class participation was 100 percent and student participation was 95 percent. The reported overall participation rate was 64 percent after replacement. IEA also applied a population-coverage qualification to the U.S. results.
Reading Achievement Scale
PIRLS reading achievement is reported on a 0–1,000 scale. The scale centerpoint of 500 was established from the 2001 achievement distribution and is maintained across assessment cycles. The reading achievement scale and the Students Like Reading context scale are separate measures and should not be interpreted as interchangeable scores.
Interpretation Limit
The table describes an association between reported reading attitudes and reading achievement. It does not establish that liking reading changes achievement, or that achievement determines how much students like reading. Differences among the three U.S. achievement means are described numerically unless an official statistical comparison is available.