Why Focusing on Test Metrics May Impede Gender Equity: Policy Insights

Abstract

Accountability policies incentivize school districts to close achievement gaps on standardized tests of math and reading, but these policies omit one prominent student dimension: gender. According to the “mixed” evidence on male–female gender gaps in math test scores, a gap advantaging males may be opening early in elementary school, but the dominant evidence—standardized testing—does not reveal this emergence. In contrast to math, the evidence for the reading gender gap favoring females is clearer, but there too the apparent female advantage may not be as large as it seems. Looking across well-established large-scale tests in math and reading, this article looks to explain why some gender gaps emerge and how policymakers can help mitigate the gaps. One of the most consistent predictors of gaps in both math and reading is gender bias. Focusing on gender gaps in tests is counterproductive to actual gender equity in education, which will require a much stronger focus on uncovering and addressing gender bias.

Keywords

gender achievement gaps tests education policy gender bias

Accountability policies may be right to exclude gender as a subgroup, but it is wrong for policymakers and educators to ignore the sources of gender bias linked to test-based gender gaps.

Key Points

Accountability policies in the United States often use subgroup comparisons (e.g., White students and Black students) on math and reading state tests to assess progress toward closing achievement gaps, but these policies do not specify gender comparisons.

State tests in math tend to reveal little to no gender difference, but many other tests find a male advantage in math and in math confidence, which both feed achievement gaps.

In reading/English language arts, boys lag behind girls on all test measures, especially attitudes toward reading.

Gender bias consistently predicts math gender gaps favoring boys, may contribute to reading gaps favoring girls, and predicts gender differences in higher education and beyond.

Policies aiming for gender equity may be more effective if they focus on gender bias rather than on test-based gender differences.

Introduction

Opportunity gaps, degree-attainment gaps, and test-based achievement gaps receive considerable interest in education research and policymaking, with test-based achievement gaps receiving some of the most prominent attention. Gaps by race and socioeconomic status (SES) tend to be among the largest measured gaps (Reardon, 2011; Reardon, Robinson-Cimpian, & Weathers, 2015) and, thus, are often the primary foci of accountability policies aimed at measuring, documenting, and reducing the gaps. The accountability era has emphasized standardized testing achievement gaps—albeit with questionable effectiveness (Lee, 2008). But accountability policy almost entirely excludes one dimension: gender.

Given the gender imbalance across college majors and career fields (particularly, in science, technology, engineering, and mathematics [STEM]), coupled with test-based accountability attempting to reduce early disparities in math and reading, should gender be included as a category for accountability? And if so, would test-based accountability focused on gender reduce downstream gender disparities in college majors and career choices?

The recent literature suggests the answers: “No, not with the current tests used for accountability,” due to various factors—test attributes, academic engagement, gendered beliefs about academic success, and test scores’ limits in predicting gender differentials in longer term outcomes. Moreover, national and international tests suggest the gender gap favoring males in math is likely even larger than measured on state tests; correlational studies, quasi-experiments, and experiments suggest that the gender gap in math widens partly because of gendered beliefs from society and teachers. Although the sources of the gender gap in reading are less studied, the (admittedly weaker) evidence there also suggests that gender stereotypes may play a role in the gap development, in this case suppressing males’ achievement.

Each section reviews the evidence for or against a gender advantage on test-based metrics. I first describe the average gaps, then variation in the gaps. In some cases—math, in particular—the test-based evidence leaves a mixed impression as to the direction and size of the gender gap; the most likely explanations for discrepancies help account for why different tests yield different results. In the most consistent findings on sources of the gender gap across the test, the overall patterns suggest that test-based gender gaps receive too much attention; they may obscure or even protect gender stereotyping. It would likely be more productive for researchers and policymakers to focus on addressing the sources of gender bias that present obstacles to both boys and girls.

Before proceeding, note that this article discusses gender gaps with a binary classification for “boys” and “girls” because almost no data addresses the achievement of transgender, gender nonbinary, or gender nonconforming students. Future state, national, and international data collection efforts should represent transgender+ students.

Math and Gender

Now, I don’t even understand why you’re looking at girls’ math achievement. These are my students’ standardized test scores, and there are absolutely no gender differences. See, the girls can do just as well as the boys if they work hard enough. (Elementary school teacher, Illinois)

Evidence of Test-Based Gaps

Mean achievement differences

Superficially, local, state, national, and international assessments provide a mixed account as to whether boys or girls hold a relative advantage in math, but the overall pattern suggests boys do. Furthermore, this advantage is probably larger than it appears on the most commonly used assessments.

State standardized tests, the most commonly used, show little evidence of gender achievement gaps in math. In 10 states’ standardized tests, “For grades 2 to 11, the general population no longer shows a gender difference in math skills” (Hyde, Lindberg, Linn, Ellis, & Williams, 2008, p. 495). Similarly, a new study using all states’ test data finds that the gender gap in math for Grades 3 to 8 is very small, at 0.03 standard deviations (SDs) in males’ favor (Reardon, Fahle, Kalogrides, Podolsky, & Zárate, 2019). (Note: gender differences presented in SD units give a sense of how different male and female achievement is, relative to how variable achievement is overall; magnitudes up to 0.2 SDs are often considered “small” in terms of practical significance, though in education [rather than experimental psychology] some have argued that even these “small” differences may be of greater importance.)

Somewhat larger male-favoring math gender gaps appear in the National Assessment of Education Progress (NAEP), a cross-sectional nationally representative assessment. The 2017 NAEP shows a male advantage of 0.07 SDs in Grade 4, narrowing to about 0.02 SDs in Grade 8, before widening again to 0.08 SDs in Grade 12. (The pattern of gap narrowing during the middle school years is consistent across other data sets, including state tests [Reardon et al., 2019] and Early Childhood Longitudinal Study–Kindergarten Class [ECLS-K; Robinson & Lubienski, 2011].) Although relatively small, the NAEP average gaps are larger than the gaps on state tests. In almost every state, the NAEP gap is larger than the state-test gap (Reardon, Kalogrides, Fahle, Podolsky, & Zárate, 2016).

Even larger gender gaps favoring males appear in the nationally representative ECLS-K of 1998-1999. The ECLS-K offered the first—and still only—opportunity to observe the gender gap in math when students began kindergarten, several years before state-mandated testing begins, and then follow them through Grade 8. The ECLS-K revealed no math gender gap when students began kindergarten, but a male-favoring gender gap of 0.25 SDs by Grade 3 (Robinson & Lubienski, 2011).

Variance and distributions

Average gender gaps, however, mask larger differences throughout the achievement distribution. First, males generally vary in math achievement scores more than females (Cimpian, Lubienski, Timmer, Makowski, & Miller, 2016; Hedges & Nowell, 1995; Hyde et al., 2008; Lubienski, McGraw, & Strutchens, 2004; Reardon et al., 2019). Regardless of test (e.g., state tests showing no gap or ECLS-K showing a large gap), male and female variances differ enough that males outperform females at the top, and females equal or outperform males at the bottom of the distribution.

Both cohorts of the ECLS-K (1998-1999, 2010-2011) detected male-favoring gaps at the top of the distribution in the fall of kindergarten, which spread during the first few years of elementary school (Cimpian et al., 2016; for distributional analyses with just the 1998-1999 data, see also Husain & Millimet, 2009; Robinson & Lubienski, 2011). The male advantage throughout the distribution particularly concentrates among White students, but all race groups experience math gender-gap growth throughout the achievement distribution (Husain & Millimet, 2009).

The NAEP data sets begin in Grade 4 and also show a gap favoring males at most points in the distribution, especially at the top, and a gender gap reversal at the bottom; this general pattern is consistent with the ECLS-K data sets, as well as with many other large-scale data sets showing a male advantage at higher percentiles. In the 2017 NAEP, the 10th percentile of boys score 0.05 SDs lower than the same percentile of girls in Grade 4. Yet, males outperform females at the top of the distribution: The same NAEP data show that males at the 90th percentile score 0.16 SDs higher than females at the 90th percentile in Grade 4. The gaps do not change much through Grade 12, except that boys are no longer performing worse than girls at the bottom.

Why Do Test-Based Math Gaps Differ?

State tests differ from other tests in their difficulty level, and they can function as “minimal competency tests” (Halpern-Manners, Warren, & Grodsky, 2015; Linn, 2000).

For most states and most grade levels, none of the items were at levels 3 [strategic thinking] or 4 [extended thinking]. Therefore, it was impossible to determine whether there was a gender difference in performance at levels 3 and 4. . . . An unexpected finding was that state assessments designed to meet [No Child Left Behind] requirements fail to test complex problem solving of the kind needed for success in STEM careers, a lacuna that should be fixed. (Hyde et al., 2008, p. 495)

As such, these state tests can suffer from ceiling effects, where students routinely score the maximum points, or do not assess complex problem solving. By contrast, the ECLS-K tests are adaptive, homing in on a student’s ability regardless of grade level; consequently, ceiling effects on the ECLS-K are rare (Pollack et al., 2005; Tourangeau et al., 2015).

Thus, gender differences in math may be compressed on state tests but more readily observed on tests with a wider range. Because the gaps tend to be larger at the top end of the ECLS-K distribution (as in NAEP), this may be a partial explanation; however, it does not explain gender gaps throughout the achievement distribution on the ECLS-K.

A second factor—testing-instruction interplay—may also explain the discrepancy. First, the tests differ in intent: State assessments aim to measure proficiency in state-specific standards (though the alignment between the tests, standards, and instruction may be imperfect; Polikoff, Porter, & Smithson, 2011); however, NAEP and ECLS-K math assessments are based on broader curriculum standards from the National Council of Teachers of Mathematics, a review of state and national standards, and advice from teachers and experts (Pollack et al., 2005). Second, because state tests link to standards via instruction, state tests allow teachers to teach to the test (Jennings & Bearak, 2014; Linn, 2000). Third, teachers report that girls demonstrate better learning behaviors and pay better attention to classroom instruction (Cimpian et al., 2016; DiPrete & Jennings, 2012; Robinson-Cimpian, Lubienski, Ganley, & Copur-Gencturk, 2014); boys are more likely to use—and encouraged to use—more inventive strategies when solving math problems, whereas girls are more likely to rely on strategies taught in class (Gallagher, Levin, & Cahalan, 2002; Lubienski, Makowski, & Miller, 2018). Altogether, the state tests may not reveal gender gaps in math because they test specific standards, teachers align their instruction, and girls follow that instruction better than do boys. In contrast, assessments like NAEP and ECLS-K draw on a broader range of items and skills and may require bolder problem solving, which girls are not often encouraged to engage, giving boys an opportunity to demonstrate their math skills on non-state assessments.

Two additional factors may contribute to gender gap variation across tests and throughout the distribution: the test-item format and differential effort. First, boys may perform better on multiple-choice items (relative to constructed-response items, Reardon, Kalogrides, Fahle, Podolsky, & Zárate, 2018). Second, despite an apparent advantage with multiple-choice items, boys are also more likely to answer them rapidly, which may indicate less effort (Soland, 2018). Correcting for their diminished effort, the math gap could be approximately 0.10 SDs larger than the observed gap, which suppresses the gender gap due to boys’ lower effort.

Evidence for Sources of Gender Disparities in Math

Cross-national comparative evidence

Substantial variation across countries in the math gender gap—favoring males in some countries and females in others—could call into question genetic differences in math ability (e.g., Dickerson, McIntosh, & Valente, 2015; Else-Quest, Hyde, & Linn, 2010; Guiso, Monte, Sapienza, & Zingales, 2008). Besides variation in the gender gap across countries, that variation links to between-country differences in gender equality (Else-Quest et al., 2010; Guiso et al., 2008; but see Stoet & Geary, 2015), their socioeconomic equality more broadly (Breda, Jouini, & Napp, 2018), and how much inhabitants associated males with science and females with liberal arts (Nosek et al., 2009).

Studies in the United States

Gender stereotypes exhibited by U.S. adults continue to predict gender gaps among boys and girls, relationships found with all types of data discussed above—state standardized tests, NAEP data, and other nationally representative data sets (e.g., ECLS-K). Gendered beliefs among adults relate to larger male–female gaps in math. Variation across states in explicitly stated gendered beliefs of men as income-earners and women as homemakers predicted a higher concentration of males among the top math test-scorers in NAEP data (Pope & Sydnor, 2010). In state-test data, the math gender gap favoring boys is larger in more affluent school districts; moreover, above and beyond the relationship between average school district SES and gender gaps, the larger the socioeconomic difference between adult males and females in the school district, the more the math gap tended to favor boys (Reardon et al., 2019).

Using the ECLS-K data sets to focus on teacher expectations affecting gender achievement gaps, when teachers faced a boy and a girl of the same race and SES that performed equally well on all past and present externally administered math tests and that the teacher rated equally well behaving and engaged with school, the teacher still rated the boy as more mathematically able (Robinson-Cimpian et al., 2014; replicated in Cimpian et al., 2016). That is, for a girl to be rated as mathematically capable as her fellow boy classmate, she not only needed to perform as well as him on all external tests but also be seen as working harder than him. Notably, the early elementary school teaching force is overwhelmingly female, so these female teachers are underrating their girls’ math abilities (the few male teachers did not). Female teachers’ underrating their female students from kindergarten through third grade accounts for about half of the growth in the gender achievement gap in math. In other words, if (female) teachers did not underestimate their female students, the gender gap in math might be substantially smaller.

The gender gap in math confidence may be larger than the gender gap in test scores (Else-Quest et al., 2010; Ganley & Lubienski, 2016; Parker, Van Zanden, & Parker, 2018), and the confidence-achievement relationship appears reciprocal—the more boys gain in achievement over girls, the more confident they become (Ganley & Lubienski, 2016). Thus, when teachers underestimate girls relative to boys (Cimpian et al., 2016; Lavy & Sand, 2018; Robinson-Cimpian et al., 2014) or attribute girls’ successes to hard work and boys’ successes to innate ability (Li, 1999; Tiedemann, 2000), they likely affect the confidence gap as well as the test gap.

Gendered treatment predicts part of the gender gap both throughout the distribution and at the top. The gender gap at the top of the achievement distribution links to higher SES families engaging their preschool children in more gendered activities (more dance lessons for girls, more organized sports for boys; Lubienski, Robinson, Crane, & Ganley, 2013), gendered beliefs about the role of women in the home and about math being for boys (Pope & Sydnor, 2010), more affluent families in general (Reardon et al., 2019), more male–female socioeconomic inequality among adults (Reardon et al., 2019), and more general socioeconomic inequality in society (Breda et al., 2018). That is, inequality (gendered or otherwise) or the means to engage children in gender-reinforcing activities may be exacerbating gender gaps, particularly at the top of the distribution. In addition, in two nationally representative data sets, teachers underestimated the math abilities of girls throughout the achievement distribution relative to equally performing and behaving boys—teachers not only overestimated the math abilities of high-achieving males relative to similar females but also overestimated average-achieving and low-achieving males relative to their same-ability female peers (Cimpian et al., 2016).

Evidence-Based Recommendations and Takeaways for Policymakers Regarding Math and Gender

Tests vary in item formats, content coverage, difficulty, and the effort students place on them, all of which can affect the direction and size of the observed math gender gap.

Regardless of test, boys at the top of the achievement distribution tend to outperform girls, with a slight reversal or no difference at the bottom of the distribution.

Gender gaps in math have not changed much over the past couple decades.

The more gender unequal a country, state, or school district is among adults, the more likely are math gender gaps among children, suggesting social construction of math gender gaps.

Principals and school boards should implement professional development that confronts the assumptions teachers make about innate abilities. Assumptions about why one gender may succeed at math can have serious consequences for the development of the gender gap.

Gender gaps in confidence in math are often larger than test-based gaps, and these differences in confidence may also stem from differential societal expectations.

State testmakers should include more challenging math items, and teachers should encourage students of all genders to engage in creative, bold problem solving.

Reading and Gender

Evidence of Test-Based Gaps

Mean achievement differences

In contrast to the almost nonexistent math gender gap on state tests, state standardized tests show that the average reading gender gap in public-school students in Grades 3 to 8 favors females by about 0.23 SDs (Reardon et al., 2019). Moreover, also in sharp contrast to the math gender gap, where females’ average test performance is higher than males in roughly half of the school districts, the female performance advantage in reading is systemic: “In no district [i.e., 0 out of 9,679 school districts] is males’ average performance [on reading/English language arts tests] higher than that of females” (Reardon et al., 2019, p. 16).

The most recent NAEP results show the reading gender gap favors females by approximately 0.16 SDs in Grade 4, increasing to 0.28 SDs in Grade 8, and remaining close to that level, at 0.25 SDs in Grade 12. A recent study of three decades of NAEP data also found that, across all cohorts, the gender gap tends to widen as grade level increases; in addition, these differences and the trajectories were consistent across the decades (Reilly, Neumann, & Andrews, 2018; see also Hedges & Nowell, 1995).

Although the ECLS-K revealed no math gender gap on average at the start of kindergarten, it revealed a sizable reading gender gap, already at about 0.20 SDs (Robinson & Lubienski, 2011). This gap shrinks to about 0.17 SDs in Grade 3 and 0.13 SDs in Grade 5, which puts the ECLS-K estimates very close to the current NAEP estimates. The gap in the ECLS-K widened again to about 0.21 SDs by the spring of Grade 8, mirroring the NAEP trend but in a smaller magnitude.

In addition, the math gender gap and the reading gender gap tend to be correlated, such that places where girls have a substantial advantage in reading scores, they tend to have an advantage over boys in math scores. These correlations exist across school districts in state assessments (Reardon et al., 2019) and across countries in international assessments (Rodríguez-Planas & Nollenberger, 2018).

Variance and distributions

Just as with math gender gaps, mean reading gender gaps conceal differences in gender variance and in gender differences throughout the distribution. Males tend to display greater variance in reading achievement, as demonstrated in state tests (Reardon et al., 2019) and ECLS-K (Cimpian et al., 2016; Robinson & Lubienski, 2011). In twin studies including students with and without diagnosed reading difficulties, males also vary more (Hawke et al., 2009). Large-scale and meta-analytic studies consistently find greater variability in test scores among males than females (e.g., Cimpian et al., 2016; Hedges & Nowell, 1995; Reardon et al., 2019; Reilly et al., 2018; Robinson & Lubienski, 2011).

The gender reading gap is largest at the bottom of the achievement distribution and narrows (but does not close) toward the top. For instance, while the average gap in Grade 4 is 0.16 SDs in the most recent NAEP, the gap at the 10th percentile is 0.28 SDs, and the gap at the 90th percentile is 0.09 SDs. The ECLS-K data suggest that while the reading gender gap is pervasive throughout the achievement distribution at kindergarten entry, the gap widens more at the bottom of the distribution during the early years of schooling and persists through Grade 8—generally consistent with NAEP. And the gap does not differ significantly across race groups (Husain & Millimet, 2009; Robinson & Lubienski, 2011).

Why Do Test-Based Reading Gaps Differ?

Differences in reading gap estimates may not seem as variable as those in math because the gross conclusion is the same (i.e., girls outperform boys), but the estimates do differ somewhat across tests. Test format appears to be a larger predictor of gender differences in reading than in math. Tests with more constructed-response items reliably have larger gender gaps favoring females than do tests with a higher proportion of multiple-choice items (Reardon et al., 2018; Schwabe, McElvany, & Trendtel, 2015). For example, Reardon et al. (2018) compared gender gaps in state tests (where states vary in the proportion of constructed-response items) and NAEP tests (where the proportion of constructed-response items is fixed across all states), and found that increasing the proportion of constructed-response from zero items to half of the items (which is about the range of variation across states) is associated with a 0.11 SDs increase in the female English language arts advantage in Grade 4 and an increase of 0.18 SDs in Grade 8.

Evidence for Sources of Gender Disparities in Reading

Although girls tend to outperform boys in reading, larger gender gaps in attitudes toward reading often predict the reading gender gap. For example, the reading motivation gender gap was roughly 2.5 times the reading achievement gender gap (PIRLS and PISA samples in Schwabe et al., 2015). Most studies find motivation gender gaps; however, some smaller scale studies failed to find significant motivation differences, though they found differences in other aspects like the importance of reading (e.g., Meece & Miller, 1999; Wigfield & Guthrie, 1997). Other studies have probed to see whether the distinction is in motivation alone, or in other factors, such as confidence in one’s reading ability, the value one places on reading, or the frequency of reading. Gender gaps in positive attitudes toward recreational reading were present in a large-scale U.S. study at Grade 1, but were even larger with each grade level through Grade 6; conversely, the gender gap in attitudes toward academic reading did not vary by grade level, but rather dropped uniformly regardless of gender (McKenna, Kear, & Ellsworth, 1995; for another large-scale study, see, for example, Sainsbury & Schagen, 2004).

Unlike the confidence gap in math, boys and girls apparently have similar confidence about themselves as readers (but see Wigfield & Guthrie, 1997, who found lower self-efficacy among boys); the differences appear to lie in motivation to read and the value they place on reading, with girls reporting greater motivation and value (Marinak & Gambrell, 2010; Wigfield & Guthrie, 1997). Female advantages in motivation, self-efficacy, and involvement in reading were partially predicted by how strongly the student identified with traits considered feminine (e.g., compassion) or masculine (e.g., competitiveness) (McGeown, Goodwin, Henderson, & Wright, 2012). Both feminine and masculine traits were positively correlated with all motivational dimensions (i.e., traits such as competitiveness predict intrinsic motivation to read and curiosity, and so do traits like compassion); however, the feminine traits tended to have stronger correlations. The relationships between the motivation dimensions and masculinity or femininity held even after statistically controlling for whether the student was a biological male or female, suggesting that the degree to which a student reflects gendered traits may be an independent factor in their motivation to read.

In large-scale data sets, gendered beliefs link to reading gaps favoring girls toward the top of the distribution at a high-level of analysis (states and geographic clusters of states; Pope & Sydnor, 2010). But more detailed data at the school district (Reardon et al., 2019) and classroom levels (Cimpian et al., 2016; Robinson-Cimpian et al., 2014) failed to find such evidence. The evidence relating gendered stereotypes and bias to math gender gaps is more robust.

One final contributor to the reading achievement gender gap is effort. Males put forth less effort (rapid guessing) not only on math tests but also on reading tests (Soland, 2018). Across all race-based and gender-based comparisons in both math and reading, the largest occurred for the male–female difference in reading, suggesting effort is a prominent factor in observed reading gender gaps. The unadjusted gender gap in reading appears to favor females by about 0.23 SDs in Grade 9, but much of that gap is due to males’ reduced effort; accounting for the effort differential reduced the female advantage to about 0.10 SDs (Soland, 2018). The true reading test gap may be smaller than the observed reading test gap.

Evidence-Based Recommendations and Takeaways for Policymakers Regarding Reading and Gender

Schools are not making much progress closing the reading gap, which favors girls and is present when entering kindergarten.

The gender gap is growing more pronounced among struggling readers.

Testmakers and educators should recognize that different tests of reading achievement can yield different estimates of the gender gap, due in part to item format and differential effort.

Teachers should try to boost boys’ motivation and reading for enjoyment, which often differs more than measured achievement.

It may be important for teachers—and society—to not stereotype reading as feminine.

Some evidence suggests that gendered attitudes predict reading gender gaps, but other work does not corroborate this finding and also finds no evidence of teacher bias in reading.

Conclusion

Much of the concern about K-12 math and reading test-based gender gaps assumes those earlier gaps affect longer term outcomes, such as career choices. Although prior achievement in a domain predicts entry into related majors and careers, subject-specific test-based differences between males and females during K-12 have little predictive relationship to higher education gender gaps (e.g., Riegle-Crumb, King, Grodsky, & Muller, 2012), especially when compared with measures of gender bias in higher education (Ganley, George, Cimpian, & Makowski, 2018; Leslie, Cimpian, Meyer, & Freeland, 2015). This does not mean that K-12 experiences are inconsequential for these later gender differences, perhaps most notably because the early and continual reinforcement and exacerbation of gender differences students receive during the K-12 years shape their attitudes toward different domains (e.g., Cimpian et al., 2016; Ganley & Lubienski, 2016; Li, 1999; Robinson-Cimpian et al., 2014; Tiedemann, 2000); however, it does suggest that the gender test gaps themselves do not explain much of these differences.

At each stage of education, students confront expectations based solely on their gender. The problem with a focus on test scores is that merely looking for differences in test scores without exploring the underlying sources of those differences permits (a) ignoring gendered attributions of success (e.g., teachers attributing girls’ success to hard work and boys’ success to ability) or (b) supporting notions of gender differences (e.g., reinforcing beliefs that math is for boys and reading is for girls). However, both factors themselves link to larger gender gaps in classrooms, in national data sets, and in international data sets (e.g., Cimpian et al., 2016; Nosek et al., 2009; Tiedemann, 2000). Thus, the emphasis on test-based metrics is outsized. Policymakers should look beyond these often-used accountability metrics for improving overall gender equity.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Breda

Jouini

Napp

(2018). Societal inequalities amplify gender gaps in math. Science, 359, 1219-1220.

Cimpian

J. R.

Lubienski

S. T.

Timmer

J. D.

Makowski

M. B.

Miller

E. K.

(2016). Have gender gaps in math closed? Achievement, teacher perceptions, and learning behaviors across two ECLS-K cohorts. AERA Open, 2(4), 1-19.

Dickerson

McIntosh

Valente

(2015). Do the maths: An analysis of the gender gap in mathematics in Africa. Economics of Education Review, 46, 1-22.

DiPrete

T. A.

Jennings

J. L.

(2012). Social and behavioral skills and the gender gap in early educational achievement. Social Science Research, 41, 1-15.

Else-Quest

N. M.

Hyde

J. S.

Linn

M. C.

(2010). Cross-national patterns of gender differences in mathematics: A meta-analysis. Psychological Bulletin, 136, 103-127.

Gallagher

Levin

Cahalan

(2002). Cognitive patterns of gender differences on mathematics admissions tests. ETS Research Report Series, 2002, i-30.

Ganley

C. M.

Lubienski

S. T.

(2016). Mathematics confidence, interest, and performance: Examining gender patterns and reciprocal relations. Learning and Individual Differences, 47, 182-193.

Guiso

Monte

Sapienza

Zingales

(2008). Culture, gender, and math. Science, 320, 1164-1165.

Halpern-Manners

Warren

J. R.

Grodsky

(2015). High-stakes testing and the rise of the GED. The ANNALS of the American Academy of Political and Social Science, 663,292-330.

10.

Hawke

J. L.

Olson

R. K.

Willcut

E. G.

Wadsworth

S. J.

DeFries

J. C.

(2009). Gender ratios for reading difficulties. Dyslexia, 15(3), 239-242.

11.

Hedges

L. V.

Nowell

(1995). Sex differences in mental test scores, variability, and numbers of high-scoring individuals. Science, 269, 41-45.

12.

Husain

Millimet

D. L.

(2009). The mythical “boy crisis”? Economics of Education Review, 28, 38-48.

13.

Hyde

J. S.

Lindberg

S. M.

Linn

M. C.

Ellis

A. B.

Williams

C. C.

(2008). Gender similarities characterize math performance. Science, 321, 494-495.

14.

Jennings

J. L.

Bearak

J. M.

(2014). “Teaching to the test” in the NCLB era: How test predictability affects our understanding of student performance. Educational Researcher, 43, 381-389.

15.

Lavy

Sand

(2018). On the origins of gender gaps in human capital: Short- and long-term consequences of teachers’ biases. Journal of Public Economics, 167, 263-279.

16.

Lee

(2008). Is test-driven external accountability effective? Synthesizing the evidence from cross-state causal-comparative and correlational studies. Review of Educational Research, 78, 608-644.

17.

Leslie

S. J.

Cimpian

Meyer

Freeland

(2015). Expectations of brilliance underlie gender distributions across academic disciplines. Science, 347, 262-265.

18.

(1999). Teachers’ beliefs and gender differences in mathematics: A review. Educational Research, 41, 63-76.

19.

Linn

R. L.

(2000). Assessments and accountability. Educational Researcher, 29, 4-16.

20.

Lubienski

S. T.

Makowski

Miller

(2018, April). “Bold problem solving”: A new construct in the pursuit of gender equity. Paper presented at the annual meeting of the American Educational Research Association, New York, NY.

21.

Lubienski

S. T.

McGraw

Strutchens

(2004). NAEP findings regarding gender: Mathematics achievement, student affect, and learning practices. In Kloosterman

Lester, Jr.

F. K.

(Eds.), Results and interpretations of the 1990 through 2000 mathematics assessments of the national assessment of educational progress (pp. 305-336). Reston, VA: National Council of Teachers of Mathematics.

22.

Lubienski

S. T.

Robinson

J. P.

Crane

C. C.

Ganley

C. M.

(2013). Girls’ and boys’ mathematics achievement, affect and experiences: Findings from ECLS-K. Journal for Research in Mathematics Education, 44, 634-645.

23.

Marinak

B. A.

Gambrell

L. B.

(2010). Reading motivation: Exploring the elementary gender gap. Literacy Research and Instruction, 49, 129-141.

24.

McGeown

Goodwin

Henderson

Wright

(2012). Gender differences in reading motivation: Does sex or gender identity provide a better account? Journal of Research in Reading, 35, 328-336.

25.

McKenna

M. C.

Kear

D. J.

Ellsworth

R. A.

(1995). Children’s attitudes toward reading: A national survey. Reading Research Quarterly, 30, 934-956.

26.

Meece

J. L.

Miller

S. D.

(1999). Changes in elementary school children’s achievement goals for reading and writing: Results of a longitudinal and an intervention study. Scientific Studies of Reading, 3, 207-229.

27.

Nosek

B. A.

Smyth

F. L.

Sriram

Lindner

N. M.

Devos

Ayala

. . . Greenwald

A. G.

(2009). National differences in gender-science stereotypes predict national sex differences in science and math achievement. Proceedings of the National Academy of Sciences of the United States of America, 106, 10593-10597.

28.

Parker

P. D.

Van Zanden

Parker

R. B.

(2018). Girls get smart, boys get smug: Historical changes in gender differences in math, literacy, and academic social comparison and achievement. Learning and Instruction, 54, 125-137.

29.

Polikoff

M. S.

Porter

A. C.

Smithson

(2011). How well aligned are state assessments of student achievement with state content standards? American Educational Research Journal, 48, 965-995.

30.

Pollack

J. M.

Rock

D. A.

Weiss

M. J.

Atkins-Burnett

Tourangeau

West

Hausken

E. G.

(2005). Early Childhood Longitudinal Study, Kindergarten Class of 1998-99 (ECLS-K): Psychometric report for the third grade (NCES 2005-062). Washington, DC: National Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education.

31.

Pope

D. G.

Sydnor

J. R.

(2010). Geographic variation in the gender differences in test scores. Journal of Economic Perspectives, 24, 95-108.

32.

Reardon

S. F.

(2011). The widening academic achievement gap between the rich and the poor: New evidence and possible explanations. In Duncan

G. J.

Murnane

R. J.

(Eds.), Whither opportunity (pp. 91-116). New York, NY: Russell Sage Foundation.

33.

Reardon

S. F.

Robinson-Cimpian

J. P.

Weathers

E. S.

(2015). Patterns and trends in racial/ethnic and socioeconomic academic achievement gaps. In Ladd

Goertz

(Eds.) Handbook of research in education finance and policy, 2nd Edition (pp. 491-509). Mahwah, NJ: Erlbaum.

34.

Reardon

S. F.

Fahle

E. M.

Kalogrides

Podolsky

Zárate

R. C.

(2019). Gender achievement gaps in us school districts. American Educational Research Journal. Advance online publication. doi:10.3102/0002831219843824

35.

Reardon

S. F.

Kalogrides

Fahle

E. M.

Podolsky

Zárate

R. C.

(2016, March). Test format and the variation of gender achievement gaps within the United States. Paper presented at the annual conference of the Society for Research on Educational Effectiveness, Washington, DC.

36.

Reardon

S. F.

Kalogrides

Fahle

E. M.

Podolsky

Zárate

R. C.

(2018). The relationship between test item format and gender achievement gaps on math and ELA tests in fourth and eighth grades. Educational Researcher, 47,284-294.

37.

Reilly

Neumann

D. L.

Andrews

(2019). Gender differences in reading and writing achievement: Evidence from the National Assessment of Educational Progress (NAEP). American Psychologist, 74(4), 445-458.

38.

Riegle-Crumb

King

Grodsky

Muller

(2012). The more things change, the more they stay the same? Prior achievement fails to explain gender inequality in entry into STEM college majors over time. American Educational Research Journal, 49, 1048-1073.

39.

Robinson

J. P.

Lubienski

S. T.

(2011). The development of gender achievement gaps in mathematics and reading during elementary and middle school: Examining direct cognitive assessments and teacher ratings. American Educational Research Journal, 48, 268-302.

40.

Robinson-Cimpian

J. P.

Lubienski

S. T.

Ganley

C. M.

Copur-Gencturk

(2014). Teachers’ gender-stereotypical ratings of mathematics proficiency may exacerbate early gender achievement gaps. Developmental Psychology, 50, 1262-1281.

41.

Rodríguez-Planas

Nollenberger

(2018). Let the girls learn! It is not only about math . . . it’s about gender social norms. Economics of Education Review, 62, 230-253.

42.

Sainsbury

Schagen

(2004). Attitudes to reading at ages nine and eleven. Journal of Research in Reading, 27, 373-386.

43.

Schwabe

McElvany

Trendtel

(2015). The school age gender gap in reading achievement: Examining the influences of item format and intrinsic reading motivation. Reading Research Quarterly, 50, 219-232.

44.

Soland

(2018). Are achievement gap estimates biased by differential student test effort? Putting an important policy metric to the test. Teachers College Record, 120(12), Article n12.

45.

Stoet

Geary

D. C.

(2015). Sex differences in academic achievement are not related to political, economic, or social equality. Intelligence, 48, 137-151.

46.

Tiedemann

(2000). Gender-related beliefs of teachers in elementary school mathematics. Educational Studies in Mathematics, 41, 191-207.

47.

Tourangeau

Nord

Lê

Wallner-Allen

Hagedorn

M. C.

Leggitt

. . . Mulligan

G. M.

(2015). Early Childhood Longitudinal Study, Kindergarten Class of 2010-11 (ECLS-K:2011): User’s manual for the ECLS-K:2011 kindergarten-first grade data file and electronic codebook, public version (NCES 2015-078). Washington, DC: National Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education.

48.

Wigfield

Guthrie

J. T.

(1997). Relations of children’s motivation for reading to the amount and breadth of their reading. Journal of Educational Psychology, 89, 420-432.

Why Focusing on Test Metrics May Impede Gender Equity: Policy Insights

Abstract

Keywords

Tweet

Key Points

Introduction

Math and Gender

Evidence of Test-Based Gaps

Mean achievement differences

Variance and distributions

Why Do Test-Based Math Gaps Differ?

Evidence for Sources of Gender Disparities in Math

Cross-national comparative evidence

Studies in the United States

Evidence-Based Recommendations and Takeaways for Policymakers Regarding Math and Gender

Reading and Gender

Evidence of Test-Based Gaps

Mean achievement differences

Variance and distributions

Why Do Test-Based Reading Gaps Differ?

Evidence for Sources of Gender Disparities in Reading

Evidence-Based Recommendations and Takeaways for Policymakers Regarding Reading and Gender

Conclusion

Footnotes

Declaration of Conflicting Interests

Funding

References