Elsevier

Intelligence

Volume 76, September–October 2019, 101377
Intelligence

Declines in vocabulary among American adults within levels of educational attainment, 1974–2016

https://doi.org/10.1016/j.intell.2019.101377 Get rights and content

Highlights

  • When controlled for educational attainment, adults' vocabulary skills have declined.

  • The vocabulary of U.S. college graduates was lower in the 2010s vs. the late 1970s.

  • Vocabulary declined across all levels of educational attainment.

  • The decline in vocabulary is primarily a time period effect.

Abstract

We examined trends over time in vocabulary, a key component of verbal intelligence, in the nationally representative General Social Survey of U.S. adults (n = 29,912). Participants answered multiple-choice questions about the definitions of 10 specific words. When controlled for educational attainment, the vocabulary of the average U.S. adult declined between the mid-1970s and the 2010s. Vocabulary declined across all levels of educational attainment (less than high school, high school or 2-year college graduate, bachelor's or graduate degree), with the largest declines among those with a bachelor's or graduate degree. Hierarchical linear modeling analyses separating the effects of age, time period, and cohort suggest that the decline is primarily a time period effect. Increasing educational attainment has apparently not improved verbal ability among Americans. Instead, as educational attainment has increased, those at each educational level are less verbally skilled even though the vocabulary skills of the whole population are unchanged.

Introduction

Are Americans more intelligent than a few decades ago, or less intelligent? In this paper, we examine one key aspect of this question: trends in American adults' verbal ability between the 1970s and the 2010s via a measure of vocabulary skills, an important indicator of the verbal component of intelligence (e.g., Carroll, 1993) and one of the highest-loading tests of g (general intelligence; Johnson, Bouchard, Krueger, McGue, & Gottesman, 2004; Lehrl & Fischer, 1988).

A theoretical case can be made for both an increase and a decrease over time in verbal ability. In support of increasing verbal skills, studies have consistently found rising IQ scores in the general population across several countries, known as the Flynn effect (Flynn, 1987; Sundet, Barlaug, & Torjussen, 2004). If IQ is increasing, and vocabulary is an important marker of IQ, then vocabulary should increase over time. In addition, educational attainment has risen. More students finish high school, and more attend college (U.S. Census, 2016). If education increases verbal intelligence, vocabulary should rise (Baker et al., 2015; Ceci & Williams, 1997; Nisbett et al., 2012).

There are also several arguments in support of vocabulary decreasing over time. The U.S. population has continually gotten older, and aging may account for a drop in verbal ability. However, past research has reported that aging is not strong enough to account for all the variance in verbal ability (Alwin and McCammon, 1999, Alwin and McCammon, 2001). The ethnic composition of the U.S. has also shifted, and increases in immigrant ethnic minority populations with lower verbal skills in English may lead to lower English vocabulary on average. In addition, scores on the verbal section of the SAT have declined over this time period (College Board, 2016). However, SAT scores are an imperfect measure of trends in ability because the population of students who take the test has changed over time. More high school students now take the SAT than in the 1970s, and thus the population may include more of lesser academic ability. Finally, fewer young people now read books, magazines, and newspapers, which may have a direct suppressive effect on vocabulary skills (Bauerlein, 2006; Twenge, Martin, & Spitzberg, 2019).

In this paper, we examine scores on a vocabulary test included in the General Social Survey (GSS), a nationally representative survey administered since 1974. The test includes 10 multiple choice items, each asking the participant to define a word. The ten items were taken from the Gallup-Thorndike Verbal Intelligence Test, Form A (Thorndike, 1942). Vocabulary is highly correlated with overall IQ (Carroll, 1993; Jensen, 2001; Sattler, 2008).

Previous research examining the GSS vocabulary test over time found that scores on the vocabulary test did not increase between 1974 and 2008 (Beaujean & Sheng, 2010). However, this study did not control for educational attainment, and educational attainment has risen considerably among American adults, and thus among the GSS participants (U.S. Census, 2016).

In addition, the origin of trends over time in vocabulary are unknown. It is important to determine if any changes over time are caused by time period (a cultural change that affects people of all ages), birth cohort or generation (a cultural change that affects young people the most), or age (a change with development; Campbell et al., 2015; Schaie, 1965). Mixed-effects models based on hierarchical linear modeling (HLM) allow the separation of the three effects using a technique known as age-period-cohort analysis (APC: Yang and Land, 2006, Yang and Land, 2008; for sample papers, see, e.g., Schwadel, 2011; Twenge, Sherman, & Wells, 2017; Wilson & Abbott, 2018; Yang, 2008). One study used an APC analysis on the GSS vocabulary data through 2000 and found age effects, a small time period effect, and also a cohort effect (Yang & Land, 2006). However, that study did not examine educational attainment as a control or a moderator variable or examine data between 2002 and 2016. Studies that have controlled for education (e.g., Dorius, Alwin, & Pacheco, 2016) have not used APC analysis or examined trends within education groups.

Section snippets

Method

The GSS is a nationally representative sample of Americans over 18, collected in most years between 1972 and 2016 (N = 61,602; for the variables in the current analysis, N = 29,912). The GSS data and codebooks are available online (Smith, Davern, Freese, & Hout, 2018). As suggested by the GSS administrators, we weighted the descriptive statistics by the variable WTSSALL to make the sample nationally representative of individuals rather than households. The weighting variable primarily corrects

Results

First, we verified the assumptions that educational attainment had increased and that educational attainment is correlated with vocabulary. Both were confirmed: The average number of years of school completed rose nearly two years over this time period, from 11.83 in 1974 to 13.68 in 2016, and years of education correlated r (29,860) = 0.47, p < .001 with vocabulary.

We then turned to the question of trends in verbal intelligence over time. Americans' scores on the GSS vocabulary test did not

Discussion

When controlled for educational attainment, American adults' vocabulary (a key indicator of verbal ability) declined between the mid-1970s and the mid-2010s. The vocabulary of American college graduates declined more than a half a standard deviation over this time period, and vocabulary also declined among those without a four-year college degree.

Why might this have occurred? First, Americans' vocabularies might be shrinking despite the increase in education. This is plausible given the steep

References (39)

  • D.F. Alwin et al.

    Aging versus cohort interpretations of intercohort differences in GSS vocabulary scores

    American Sociological Review

    (1999)
  • D.F. Alwin et al.

    Aging, cohorts, and verbal ability

    The Journals of Gerontology Series B: Psychological Sciences and Social Sciences

    (2001)
  • R. Arum et al.

    Academically adrift: Limited learning on college campuses

    (2010)
  • D. Bates et al.

    lme4: Linear mixed-effects models using Eigen and S4. R package version 1

    (2014)
  • M. Bauerlein

    The dumbest generation: How the digital age stupefies young Americans and jeopardizes our future

    (2006)
  • A.A. Beaujean et al.

    Examining the Flynn effect in the General Social Survey Vocabulary test using item response theory

    Personality and Individual Differences

    (2010)
  • W.K. Campbell et al.

    Generational differences are real and useful

    Industrial and Organizational Psychology: Perspectives on Science and Practice

    (2015)
  • J.B. Carroll

    Human Cognitive Abilities: A Survey of Factor-Analytic Studies

    (1993)
  • S.J. Ceci et al.

    Schooling, intelligence, and income

    American Psychologist

    (1997)
  • Cited by (11)

    • THE FLINN EFFECT IN RUSSIA: IMPACT OF SETTLEMENTS’ SIZE

      2021, Experimental Psychology (Russia)
    View all citing articles on Scopus
    View full text