The Armed Services Vocational Aptitude Battery (ASVAB): Little more than acculturated learning (Gc)!?

The Armed Services Vocational Aptitude Battery (ASVAB) is administered to over 1 million participants in the USA each year, serving either as a screening test for military enlistees or as a guidance counseling device in high schools. In this paper, we examine the factorial composition of the ASVAB in relation to the theory of fluid and crystallized intelligence and Carroll's [1993. Human cognitive abilities: a survey of factor-analytic studies. New York: Cambridge Univ. Press.] three-stratum model. In two studies (N=349, N=6751), participants were administered both the ASVAB and tests designed to measure factors underlying these (largely) analogous models. Exploratory and confirmatory factor analyses (CFA) of correlational data suggested that the ASVAB primarily measures acculturated learning [crystallized intelligence (Gc)]. This evidence does not support the frequent claim that this test measures psychometric g. Our conclusion is that the ASVAB should be revised to incorporate the assessment of additional broad cognitive ability factors, particularly fluid intelligence and learning and memory constructs, if it is to maintain its postulated function.     

Reading and listening at three school stages: cloze tests and their response analysis

This paper reports the results of cloze tests in the reading and listening modes together with a computer analysis of responses to the tests. The subjects were groups of Scottish school children at the ages of 8-9, 11-12 and 13–14 years sampled over the whole country as part of a national survey of English language; the cloze tests were only a small part of the whole testing programme which also contained three other major reading tests. Approximately 400 subjects took cloze tests in each mode at each age. The test material was the same throughout for all stages tested. Two tests, each containing one narrative and one expository text were used. The mode of presentation did not significantly affect the types of cloze responses offered nor the total scores of the tests at any stage. However, results indicated better performance for older subjects when they read, and for the youngest group when they listened to, expository though not narrative passages. The comparisons of the results for the three different school stages showed continuing interdependence of reading and listening ability through the ages tested. The different cloze response patterns for the two types of text (in either mode) as well as the only moderate correlation between the texts, indicated that success in comprehending narratives may not necessarily transfer to comprehending information.     

CLOZE MEASURES AS INDICES OF ACHIEVEMENT COMPREHENSION WHEN LEARNING FROM EXTENDED PROSE

The hypothesis that cloze measures are a function of content achievement among adult learners and, consequently, should be sensitive to instructional treatments was tested in two experimental studies. College juniors and seniors took tests immediately before (pre), immediately after (post) and four weeks after (delay) studying a prose passage. The types of tests administered in each session were: (I) a 20-item multiple-choice test, (2) a reproduction passage cloze test, (3) a recognition passage cloze test, (4) a reproduction summary cloze test, and (5) a recognition summary cloze test. All tests show significant differences between pre- and posteonditions, and between recognition and reproduction modes. The reproduction summary cloze test was found to be the most sensitive to the instructional treatment, as indicated by an oJ 2 statistic on pre-post measures. The summary cloze tests were resistant to forgetting while the cloze passage and multiple-choice tests show significant decreases in performance over the four week delay interval.     

Metacognitive factors in scientific problem-solving strategies

The aim of this study on 42 seventh graders (ages 12–13) was to determine whether and to what extent students’ metacognitive level is linked to their conceptualization and performance in problem solving at school, especially science problems. This hypothesis is supported by a number of studies showing that metacognition is a factor in learning. Two indexes were devised for the study: an index of metaknowledge about classroom learning, and an index of metacognitive monitoring in relation to task difficulty on a non-academic problem. These two indexes were related to the students’ intelligence test scores and solving strategies on electricity problems. The results showed that (1) the metaknowledge level was more closely tied to crystallized intelligence (Gc), and (2) metacognitive monitoring was more closely associated with fluid intelligence (Gf). Moreover, both metacognitive indexes were strongly linked to scientific problem-solving strategies (correlations around .50).     

Prediction of reading by Ga and Gc specific cognitive abilities for low‐SES white and hispanic English‐speaking children

Based on previous findings (McGrew, Flanagan, Keith, & Vanderwood, 1997) that auditory processing (Ga) predicts reading decoding, and crystallized intelligence (Gc) predicts reading comprehension beyond the prediction of reading by g(broad general intelligence), this study examined the cross‐ethnic predictive validity of Ga and Gc for reading achievement among low‐SES English‐speaking White and Hispanic students. Subtests of the Woodcock‐Johnson Psycho‐Educational Battery‐Revised (WJ‐R) cognitive subtests were used to measure Ga and Gc, and the WJ‐R achievement subtests were used to measure basic reading skills and reading comprehension. Results indicated that there are no differences between ethnic groups in the prediction of reading ability, and that phonetic coding and crystallized intelligence together are strong predictors of reading achievement. © 2000 John Wiley & Sons, Inc.     

Measuring college students' reading comprehension ability using cloze tests

Recent investigations challenge the construct validity of sustained silent reading tests. Performance of two groups of post‐secondary students (e.g. struggling and non‐struggling) on a sustained silent reading test and two types of cloze test (i.e. maze and open‐ended) was compared in order to identify the test format that contributes greater variance in reading comprehension. One hundred participants were recruited from students enrolled in a preparatory course for a high‐stakes statewide reading examination. Our results suggest that all three measures have good concurrent validity. There was no evidence that open‐ended cloze performance was more related to verbal ability than any other reading measure. Maze performance did the best job at discriminating between our struggling and non‐struggling readers. Implications for reading comprehension assessment in post secondary‐aged adults are discussed.     

Self estimates of general,crystallized, and fluid intelligences in an ethnically diverse population

Self-estimated intelligence is a quick way to assess people's conceptions of their own abilities. Furnham (2001) and colleagues have used this technique to make comparisons across culture and gender and different approaches to intelligence (such as g or Multiple Intelligences). This study seeks to build on past work in two ways. First, a large, diverse sample (N = 2309) enables the study of self-estimated intelligence across ethnicity. Second, one of the most prominent accepted intelligence theories is the Cattell–Horn–Carroll (CHC) theory. Can laypeople distinguish between general, fluid, and crystallized intelligence? This study finds evidence for a “modesty bias” in Hispanic Americans. In addition, fluid and crystallized intelligences may simply be seen as general intelligence (particularly by Caucasians and males).     

Language assimilation and performance in achievement tests among Hispanic children in the U.S.: Evidence from a field experiment

We provide new evidence about the effect of testing language on test scores using data from two rounds (conducted approximately six years apart) of the New Immigrants Survey. In each round, U.S.-born and foreign-born children of Hispanic origin were randomly assigned to take the Woodcock-Johnson achievement (two reading and two math) tests, either in Spanish or in English. U.S.-born children of Hispanic immigrants perform better in reading tests (but not in math tests) when they are assigned to take tests in English. The size of the testing-language effect remains stable across rounds. Foreign-born children of Hispanic immigrants perform better in both reading and math tests when they are assigned to take tests in Spanish in the first round. However, the size of the testing-language effect declines in reading tests and completely disappears in math tests by the second round. Our results suggest that the depreciation of Spanish skills is an essential factor (and, in some cases, more important than the accumulation of English skills) in explaining the decline in the testing-language effect among foreign-born children. We also explore how age at immigration and years spent in the U.S. affect language assimilation.     

Structural extensions of a hierarchical view on human cognitive abilities

During the 1990s, research within the Individual Differences Unit at the University of Sydney was oriented towards expanding knowledge of poorly understood domains of human cognition. Using the theory of fluid and crystallized intelligence (Gf/Gc) served as the overarching framework. This research has led to an improved understanding of the nature of complexity and the role of working memory in Gf. Another outcome has been the charting of factors underlying of mental speed, tactile/kinesthetic performance, and olfactory memory. Recently, there has been exploration of the “borderline area” between abilities and personality. Our work on self-confidence and self-monitoring of cognitive performance has demonstrated excellent measurement properties of scores based on confidence ratings. We have also established a clear separation of a broad self-confidence factor from traditional intelligence and personality factors. Our research in these borderline areas has also explored the status of emotional intelligence. Results from several studies raise questions regarding the measurement of this construct.     

Characterizing mature human intelligence: Expertise development

Results from a study of 263 male players at 48 levels of expertise in the game of GO, and ranging from 18 to 78 years of age, suggest a need to revise the extended theory of fluid (Gf) and crystallized (Gc) intelligence to take account of continued development of intelligence throughout adulthood. The extended theory of Gf–Gc is based on evidence that Gf, short-term apprehension and retrieval (SAR) and cognitive speed (Gs), decline with age over adulthood. Results from a number of studies, however, suggest that within the domains of expertise, high levels of reasoning, feats of memory and speeded thinking similar to Gf are displayed by older adults. To explore this hypothesis, measures of reasoning, memory and cognitive speed were constructed within the domain of expertise related to playing the complex game of GO. Analysis of the structure of the GO-embedded measures and standard measures of Gf (SAR and Gs) indicated a form of short-term memory — labeled expertise working memory (EWM) — that had substantially wider span than the short-term working memory (STWM) of SAR. This finding is consistent with the hypothesis that long-term working memory is built up during the course of developing high levels of expertise. The results also suggest that a form of expertise deductive reasoning (EDR), utilizing EWM and incorporating large stores of knowledge, is distinct from Gf. Expertise cognitive speed (ECS), however, was not found to be reliably distinct from the Gs factor. Analyses of cross-sectional age differences indicate an age-related decline in both EDR and EWM, but as higher levels of expertise are reached, age-related decline does not occur. To the extent that there is continued press to advance expertise throughout adulthood, there may be improvement, not decline, in the EDR and EWM forms of intelligence.     

Examining the role of vocabulary depth, cross-linguistic transfer, and types of reading measures on the reading comprehension of Latino bilinguals in elementary school

Given the increase of bilingual students in the K-12 public school system, understanding reading comprehension performance, especially among this population, has been a major focal point in the research literature. This study explores the nature of reading comprehension among a sample of 123 Spanish–English bilingual elementary students. We add to the existing knowledge base regarding reading comprehension in two significant ways: (1) augmenting the Simple View of Reading by testing the role of both vocabulary depth contribution and dual-linguistic ability in English reading comprehension; and (2) questioning the manner through which reading comprehension is understood through measurement and conceptualization. Specifically, we build a comprehensive model of reading comprehension that tests the effects for vocabulary depth, Spanish oral language, and biliteracy. In line with previous research that suggests different reading measures tap different abilities, we test our model for three different measures of reading comprehension: a cloze exercise, a passage and multiple choice based test, and a timed silent sentence reading judgment task. Our findings converge with previous research on the role of vocabulary depth in reading comprehension and also challenge prior work which has compared different reading measures. Implications for theoretical and empirical approaches to understanding reading comprehension, specifically among Spanish–English bilingual students, are discussed.     

Testing reading comprehension of theoretical discourse with cloze

The ability of cloze tests containing frequent, every n‐th word deletions to measure comprehension of macropropositions has been challenged on both theoretical and empirical grounds, calling into question the validity of such tests for assessing comprehension of much of the discourse encountered by university‐level students. To evaluate the comprehension of a writer’s reasoning, it is recommended that cloze tests position gaps so as to target recognition of cohesive devices and the ability to draw inferences from other sentences. To test the validity of such a design, a large sample of scores on discourse cloze tests administered in introductory college economics is compared to scores on true–false comprehension tests designed to target recognition of connective propositions. The two distributions of scores do not differ significantly in terms of mean value, dispersion or frequency distribution, suggesting that appropriately designed cloze tests can provide a valid assessment of the reader’s integration of theoretical text. In addition, the usefulness of readability formulas based on surface characteristics of text is challenged when readability is defined in terms of the difficulty of constructing a coherent representation of theoretical text.     

Crystallized and fluid intelligence in a “new” instructional situation

Twenty-four measures of crystallized intelligence (Gc) and fluid intelligence (Gf) were obtained for samples of graduates and failures of an innovative instructional setting where computer-managed mastery learning was implemented to teach elementary electricity and electronics. Seven stepwise multiple discriminant analyses and associated statistics were computed to determine which linear combinations of Gf and Gc measures could optimally separate the two groups. Corresponding classification functions derived for the discriminant analyses were applied to these measures obtained from the subjects to evaluate the effectiveness of differentiating failures and graduates. The results did not substantiate the hypothesis that Gf measures would be associated with student success in new instructional situations more so than Gc measures. Contrary to theory, the findings suggest that some unconventional educational environments are not necessarily dysfunctional for better abled students. In these situations they can just as easily exercise and exploit those skills developed and applied in more traditional instructional settings.     

Reciprocal effects between fluid and crystallized intelligence and their dependence on parents' socioeconomic status and education

The investment theory of Cattell supposes an influence of fluid on crystallized intelligence. The development of fluid intelligence largely depends on biological factors, of crystallized intelligence on fluid intelligence and environmental stimulation. To test this theory two contrasting samples representing a broad ability range were chosen, a Brazilian sample (ages 7 to 15, N = 833, mean IQ 92) and a sample with a higher ability level in Germany (ages 11 to 19, N = 722, IQ 118). Analyses of cross-lagged effects across two year intervals show similar effects of fluid intelligence on crystallized and vice versa (around β = .17). Parental socioeconomic status and parental education have in both samples a slightly stronger effect on crystallized than on fluid intelligence. The first result refutes Cattell's theory, the second gives some support. The development of fluid intelligence also seems to be influenced by non-biological environmental factors resulting in a concept of intelligence as a malleable ability.     

Qualitative aspects of gain on achievement tests

Summary and Conclusion The basic questions involved in this investigation were the effect of initial standing and item difficulty on the gain established on final over initial measures in achievement testing. The hypotheses were advanced that the use of gain discriminated against bright students on two counts; first, by limiting possibility for gain and second, that equal gains might really represent different levels of performance. It has been shown that the use of a measure of achievement, the effectiveness index, which is independent of initial standing, shows a higher positive correlation with intelligence than the correlation obtained between gain and intelligence. The use of partial correlation to hold the factor of initial standing constant raised the relation found between gain and intelligence in 11 out of 12 cases, indicating that the higher initial standing of brighter students tended to limit their performance. Further, it has been shown that the gain of bright students tends to be established on more difficult items. In addition, the film apparently facilitated the success of students with above average intelligence on difficult items since the relationship found between intelligence and gain or nongain on test items was markedly higher in experimental than in control sections. To the extent that the gains made on the tests employed in this investigation are representative of gains made on achievement tests in general, in pretest and posttest situations, it may be concluded that bright students are discriminated against when gain is the measure of performance. The bright students must overcome the twin difficulties of establishing their gain on more difficult items in competition with students who havemore items on which improvement, i.e., gain, may be established. It is obvious that groups and individuals differing widely in their initial performance on tests should not be compared using gain as a criterion unless there is some compensation for these factors. This paper funishes additional evidence relative to a statistical procedure—the “Effectiveness Index”—discussed by Hovland, Lumsdaine, and Sheffield inExperiments on Mass Communication. It was originally presented to Section Q of the American Association for the Advancement of Science at the St. Louis Meeting, December 1952. The Study was made possible by a grant from the University of Nebraska Research Council. Dr. Herbert A. Smith is Associate Professor of Secondary Education and Supervisor of Science, the University of Nebraska.     

The effect on the doze test of changes in deletion frequency

Although the pseudo-random cloze procedure has been in use for some twenty-five years as a measure of readability and reading comprehension, little research has been carried out into the effect of deleting words from text more or less frequently. This paper reports on an experiment in which the deletion frequency variable was systematically studied. Every 6th, 8th, 10th and 12th word was removed from three texts of differing difficulty, and the effect studied. Significant differences among cloze tests resulted, but the differences were unpredictable. Deleting every 12th word did not necessarily result in an easier test than deleting every 6th 8th or 10th word. However, when only items identical to both cloze tests under consideration were compared, no significant differences were found. It appears that cloze items are, on the whole, unaffected by context greater than five words. Testers are warned that changing deletion frequency may result in a different measure of readability or comprehension.     

Confirmatory factor analysis of the KABC‐II in preschool children

The present study assessed the Kaufman Assessment Battery for Children, Second Edition (KABC‐II) in relation to the synthesized Cattell–Horn–Carroll (CHC) theory of intelligence with a preschool sample. Participants were 200 preschool children between four and five years of age. A confirmatory factor analysis (CFA) was conducted, and different variations of the CHC model were examined to determine which provided the best representation of the proposed underlying CHC constructs tested by the KABC‐II. The models included one similar to Spearman's g, a contemporary two‐stratum model consisting of fluid and crystallized intelligence (Gf‐Gc model), and a synthesized CHC broad factor +g model. The last was the empirically validated theory of interest in this study. Results of the CFA revealed that the broad factor +g CHC model was the best overall design to explain KABC‐II results. © 2009 Wiley Periodicals, Inc.     

People who cheat on tests accurately predict their performance on future tests

Studies suggest that people who cheat on a test overestimate their performance on future tests. Given that erroneous monitoring of one's own cognitive processes impairs learning and memory, this study investigated whether cheating on a test would harm monitoring accuracy on future tests. Participants had the incentive and opportunity to cheat on one (Experiments 1, 2, and 3, with N = 90, 88, and 102, respectively) or two (Experiment 4, N = 214) of four general-knowledge tests. Cheating produced overconfidence in global-level performance predictions in Experiment 2 (Cohen's d ≥ 0.35) but not in Experiments 1 or 4. Also, cheating did not affect the absolute or relative accuracy of item-level performance predictions in Experiments 3 or 4. A Bayesian meta-analysis of all experiments provided evidence against cheating-induced overconfidence in global- and item-level predictions. Overall, our results demonstrate that people who cheat on tests accurately predict their performance on future tests.     

Measuring Life-Science Concepts of Elementary-School Pupils

This investigation sought information on the performance of high and low ability Ss on two measures of creativity, and the relationship between performance on the two measures. Forty-seven 13-year-old high school students comprised the sample. High and low ability students from the top and lower than average streams were given the A.H. 4 intelligence test, the Junior-Senior High School Personality Questionnaire, and a drawing test on a specific theme, which was assessed by five expert judges for creativity on the basis of certain criteria. Significant differences were found between the high and low ability groups both in the H. S. P. Q. test and the creativity drawing test. There was also a significant correlation between the A.H. 4 test and the judges’ ranking for creativity. However, there was no significant difference between the two creativity measures, nor between the H. S. P. W. and the A.H. 4 tests. The results confirm the view that there is a relationship between intelligence and creativity, that the relationship depends on types of intelligence tests and creativity measures used, and the need exists for a definition of creativity based on commonly accepted criteria.