Gender Differences in Reading and Writing Achievement

1 Gender Differences in Reading and Writing Achievement : Evidence Fromthe National Assessment of Educational Progress (NAEP)David ReillyGriffith University, Queensland, AustraliaDavid L. Neumann and Glenda AndrewsGriffith University, Queensland, Australia, and Menzies HealthInstitute Queensland, AustraliaA frequently observed research finding is that females outperform males on tasks of verbaland language abilities, but there is considerable variability in effect sizes from sample tosample. The gold standard for evaluating Gender Differences in cognitive ability is to recruita large, demographically representative sample. We examined 3 decades of studentachievement in Reading and Writing from the National Assessment of Educational Progress todetermine the magnitude of Gender Differences (N million), and whether these weredeclining over time as claimed by Feingold (1988). Examination of effect sizes found adevelopmental progression from initially small Gender Differences in Grade 4 toward largereffects as students progress through schooling.

2 Differences for Reading were small-to-medium(d .32 by Grade 12), and medium-sized for Writing (d .55 by Grade 12) and werestable over the historical time. Additionally, there were pronounced imbalances in genderratios at the lower left and upper right tails of the ability spectrum. These results areinterpreted in the context of Hyde s (2005) Gender similarities hypothesis, which holds thatmost psychological Gender Differences are only small or trivial in size. Language and verbalabilities represent one exception to the general rule of Gender similarities, and we discuss theeducational implications of these : Gender Differences , Reading , Writing , literacy, sex differencesSupplemental materials: question of whether males and females differ incognitive abilities has been the focus of considerableresearch in recent decades. While there is a generalconsensus that males and females do not differ in generalintelligence (Halpern, 2000), Gender Differences are com-monly observed for more specific cognitive abilities suchas visual spatial ability (Voyer, Voyer, & Bryden, 1995)and language (Miller & Halpern, 2014).

3 However, Hyde(2005) had proposed the Gender similarities hypothesis(GSH), which claimed that males and females are sim-ilar on most, but not all, psychological variables. That is,men and women, as well as boys and girls, are more alikethan they are different (p. 581). It holds that most genderdifferences are small or trivial (close to zero) in magni-tude. One exception to this hypothesis may be the gendergap in Reading Achievement , which is found cross-culturally (Lynn & Mikk, 2009; Reilly, 2012) and ex-ceeds the threshold proposed by Hyde and Grabe (2008,p. 170) for nontrivial Gender difference effect sizes (d .10). In a recent review, Hyde (2014) remarked that it is difficult to reconcile (p. 382) the magnitude of thegender gap observed in Reading with other domains ofverbal ability ( , vocabulary, anagrams), which Hydeand Linn (1988) claimed are typically much the issue of Reading is received greater attention, thereis a growing body of evidence that males and females alsodiffer in Writing ability (Camarata & Woodcock, 2006; Reyn-olds, Scheiber, Hajovsky, Schwartz, & Kaufman, 2015;Scheiber, Reynolds, Hajovsky, & Kaufman, 2015).

4 Reynoldset al. (2015) noted that the issue of Gender Differences inwriting skills has been overlooked because it is less frequentlymeasured in educational assessments. In cases where writingability is assessed, researchers should examine Gender differ-ences to determine if any meaningful Differences occur. More-over, researchers should compare the size of any Differences tothose observed with Reading assessments when both domainsare examined in the same article was published Online First September 20, Reilly, School of Applied Psychology, Griffith University,Queensland, Australia; David L. Neumann and Glenda Andrews, School ofApplied Psychology, Griffith University, Queensland, Australia, and Men-zies Health Institute Queensland, concerning this article should be addressed to DavidReilly, School of Applied Psychology, Griffith University, Southport,Queensland 4222, Australia. E-mail: Psychologist 2018 American Psychological Association2019, Vol. 74, No. 4, 445 4580003-066X/19/$ researchers ( , Feingold, 1988) have claimed thatas a response to societal changes in the status and roles ofwomen, Gender Differences are declining (see the onlinesupplemental materials for a more detailed discussion ofthese issues).

5 Gender role attitudes in the United States havechanged over time, giving boys the freedom to pursuelanguage-arts fields just as an increasing number of girlsnow pursue science, technology, engineering, and mathe-matics fields. Feingold analyzed educational data from 1947to 1980, showing a decline over time. More recently, Caplanand Caplan (1997, 2016) have questioned whether genderdifferences in verbal and language abilities even existed atall and were the product of selection bias in samples, whileHyde (2005) has claimed that most Gender Differences areeither small or trivial in size. The current study examineswhether historical patterns of Gender Differences in readingand Writing are still present in modern samples and, if so, todetermine their magnitude. It presents a meta-analysis ofstudent Reading and Writing Achievement drawn from theNational Assessment of Educational Progress (NAEP), alarge nationally representative sample of students from theUnited States conducted by the National Center for Educa-tional Statistics (NCES).

6 Before turning our attention to thisdataset, we first present an overview of theoretical perspec-tives on Gender Differences in language Perspectives on Gender Differences inLanguage AbilityIn their pioneering textThe Psychology of Sex Differ-ences, Maccoby and Jacklin (1974) presented the first sys-tematic review of the psychological literature on genderdifferences, arguing that Gender Differences in verbal abilityand language were well established (p. 351) and showeda developmental progression toward larger gaps with in-creasing age. Much of the literature they reviewed focusedexclusively on Reading ability, rather than considering lan-guage proficiency more broadly with higher level tasks suchas Writing , spelling, and grammar usage. But a number ofsubsequent studies have also reported Gender differenceswith the largest being spelling and use of grammar (Reilly,Neumann, & Andrews, 2016; Stanley, Benbow, Brody,Dauber, & Lupkowski, 1992).Theoretical explanations for the emergence of genderdifferences in Reading and language proficiency have beenoffered.

7 These center around biologically based or socio-cultural explanations for Gender Differences , or combina-tions of both (Eagly & Wood, 2013; Halpern & Tan, 2001):(a) differential rates of maturation, (b) Gender Differences inlateralization of brain function, (c) Gender Differences invariability, (d) Gender Differences in externalizing behaviorand language competence, and (e) Gender -stereotyping ofreading and language as feminine traits. Each will be dis-cussed in detail Rates of MaturationGirls have a faster rate of maturation and may thereforebe attaining greater proficiency than similarly aged boys(Dwyer, 1973), making Reading easier and more an explanation holds that boys are merely delayed(developmental lag) and boys would attain an equivalentlanguage proficiency given sufficient time. However, thisclaim is inconsistent with studies showing Gender differ-ences in Reading that persist into adulthood (Kutner et al.,2007). Gender Differences in Lateralization ofBrain FunctionSome researchers have claimed that lateralization of brainfunction for language may differ between males and fe-males (Levy, 1969).

8 It has been claimed that the regionsresponsible for language tasks are strongly lateralized to theleft cerebral hemisphere in right-handed males, but thatlanguage regions in females are more likely to be distributedacross both the left and right hemisphere (B. A. Shaywitz etal., 1995). Bilateral language function presumably affordssome benefits, which could explain the female advantageobserved on such tasks. However, empirical support for theLevy hypothesis is mixed (Kaiser, Haller, Schmitz, &Nitsch, 2009), with some neuroimaging studies showinggender Differences in lateralization for language tasks (Bur-man, Bitan, & Booth, 2008; Clements et al., 2006), whileothers do not (Wallentin, 2009).David Reilly446 REILLY, NEUMANN, AND ANDREWSG ender Differences in VariabilityOne explanation for lower Reading and language profi-ciency in males is thegreater male variabilityeffect, whichstates that males show greater variability in cognitive per-formance across all cultures (Feingold, 1992; Machin &Pekkarinen, 2008).

9 Even if there were no Gender differencesin group means, the consequence of greater male variabilityis that males will be overrepresented at the extreme left tailof the ability distribution, which Hawke, Olson, Willcut,Wadsworth, and DeFries (2009) argued explains why gen-der ratios of poor readers favors females. Boys are alsooverrepresented in populations with Reading impairment,dyslexia, attention disorders, and mental retardation sug-gesting that there may be a Gender -linked neurologicalcontribution (Halpern, Beninger, & Straight, 2011). Whileexplanations for the greater male variability hypothesis inintelligence have been made by evolutionary psychologists(Geary, 2010), few specific evolutionary theories have beenproposed for verbal and language abilities (Geary, Win-egard, & Winegard, 2014), perhaps because these are morerecent in an evolutionary Differences in Externalizing Behaviorand Language CompetenceOther researchers have argued that Gender Differences inexternalizing behavior may also partly explain a greaterfemale language competence (Limbrick, Wheldall, & Mad-elaine, 2011).

10 Clinicians identify more boys than girls withexternalizing behavior and attention disorders (McGee,Prior, Williams, Smart, & Sanson, 2002) which have bothbeen associated with Reading and language impairment. Forexample, in a longitudinal sample from the United States,Rabiner and Coie (2000) reported that attention-impairmentand externalizing behavior measured in kindergarten pre-dicted later Reading impairment in fifth grade. Other studieshave followed children over longer time frames. In a lon-gitudinal study of child development in Australia, Smart,Prior, Sanson, and Oberklaid (2001) found that externaliz-ing behavior problems at age 7 predicted the severity oflater Reading and spelling difficulties at ages 13 14, evenafter controlling for intelligence and socioeconomic also present in girls, Smart et al. found that theassociation between externalizing behavior and Reading im-pairment was significantly stronger in boys. Such an asso-ciation is not necessarily causal, and may well be reciprocalin nature.