Search

Using data on students in grades 5 to 9 in the West Bank, we study the link between poor mental health and cognitive development. After controlling for a wide range of potential confounders in an entropy balancing approach, boys’ cognitive test scores are significantly associated with self- and parent-reported measures of mental health. Boys classified as having abnormal mental health scores lag about one grade level behind their peers. Among girls, however, the relationship is weaker and—depending on specification—insignificant. We also show that poor mental health, in turn, is linked with reporting exposure to potentially traumatizing events related to the Israeli–Palestinian conflict.

Israel’s consistently low ranking on international achievement tests has been exclusively attributed to insufficient investments in education or inefficient use of available resources. In contrast, this article shows this low ranking to be a true reflection of Israel’s reality in terms of characteristics that affect the mean ability of its student population to profit from schooling (MAPS). Israel’s actual ranking on the PISA test among the 34 OECD member countries (the lowest decile, in the company of Mexico and Turkey) perfectly matches its expected rank on the basis of MAPS (operationally defined as the proportion of children above the ‘poverty line’).

Background The effects of an intervention on patients from populations other than that included in a trial may vary as a result of differences in population features, treatment administration, or general setting. Determining the generalizability of a trial to a target population is important in clinical decision making at both the individual practitioner and policy-making levels. However, awareness to the challenges associated with the assessment of generalizability of trials is low and tools to facilitate such assessment are lacking. Methods We review the main factors affecting the generalizability of a clinical trial results beyond the trial population. We then propose a framework for a standardized evaluation of parameters relevant to determining the external validity of clinical trials to produce a “generalizability score”. We then apply this framework to populations of patients with heart failure included in trials, cohorts and registries to demonstrate the use of the generalizability score and its graphic representation along three dimensions: participants’ demographics, their clinical profile and intervention setting. We use the generalizability score to compare a single trial to multiple “target” clinical scenarios. Additionally, we present the generalizability score of several studies with regard to a single “target” population. Results Similarity indices vary considerably between trials and target population, but inconsistent reporting of participant characteristics limit head-to-head comparisons. Conclusion We discuss the challenges involved in performing automatic assessment of trial generalizability at scale and propose the adoption of a standard format for reporting the characteristics of trial participants to enable better interpretation of their results.

The analysis presented in this paper is part of the wider project “Determinants of Cognitive Development in Deprived Environments: Evidence from the West Bank” funded by the German Research Foundation (DFG) under grant number JU 2769/2. We are grateful to the PA Ministry of Education, test administrators, and the students who participated in the study and their parents for their time and effort

A necessary, albeit tacit assumption underlying pattern analysis of cognitive profiles is that an examinee’s profile pattern is not affected by the level of precision used in measuring the subtest, index or factor scores. We empirically test the truth of this assumption across various precision levels, such as IQ points (1/15SD), T-scores (0.1SD), scaled scores (1/3SD) and stanines (0.5SD). The results clearly refute the pattern stability assumption. They question the very uniqueness of profile patterns as a stable individual characteristic and challenge their use in both clinical practice and scientific research. Possible solutions are suggested and critically examined.

Differential granting of extra-examination time (EET) is commonly based on learning disabilities (LD) status: EET is granted to LD examinees and is denied to nondisabled examinees. We argue that LD serves as a proxy for the extent to which time limitation affects the examinee’s test score ( e). Hence, the validity of the LD-based EET granting policy depends on how well LD status serves as a proxy for e. Reanalysis of 11 comparative experimental studies of the effect of EET shows that LD status is a poor proxy for e. The proportion of nondisabled examinees who benefit from EET roughly equals the corresponding proportion among LD students. Implications of these results for the validity and fairness of this policy are discussed.

This study contributes to the investigation of the variability of the schooling effect on cognitive development between educational systems and its underlying factors, by focusing on 3 cases differing in the characteristics assumed to affect the magnitude of the schooling effect (the quality of the schooling and students’ mean ability to benefit from their schooling): the United Nations Relief and Works Agency (UNRWA) educational system in the Palestinian refugee camps in the West Bank and the 2 sectors of the Israeli educational system: the Jewish and Arab systems. Congruent with our expectations, the highest absolute and relative effects of schooling were found in the Israeli Jewish system. In contrast, the results obtained in the 2 Arab systems, namely, the higher schooling effect found in the poorer and more oppressed refugee camps of the West Bank, are surprising and require further investigation.

Schooling is now considered the major factor underlying the development of cognitive abilities. However, most studies on the effect of schooling on cognitive development have been conducted in free and generally supportive western environments. The possible variability of schooling effects between educational systems differing in the quality of schooling they provide or in the ability of the student population to benefit from schooling has been neglected. This work contributes to the study of this variability by estimating the effect of schooling on cognitive development in an educational system characterised by particularly unfavourable circumstances, which combine poor schooling quality and low ability of the student population to benefit from schooling due to economic disadvantage, military occupation, and political oppression: the UN Relief and Works Agency Palestinian refugee camps of the West Bank. The results indicate a lower contribution of schooling to children’s cognitive development, both relat...

Relying on the results of the achievement tests in mathematics, science, native language (Hebrew/Arabic) and English, administered to 1430 5th-grade co-educational classes in Israel, this study examines the between-class variability of the within-class mean score gender differences and its class and school correlates. The four main results of the study are: (1) remarkable between-test stability of the within-class gender gap; (2) considerable variability of the within-class gender gap, in terms of both sign and magnitude, found for each of the four tests: children studying in different classes are exposed to different, sometimes quite opposite, gender differences and this variability is effectively masked by the aggregate-level analyses typically reported in the literature; (3) the lion’s share of the variability of the within-class gender gap lies within, rather than between schools; and (4) the relative frequency of within-class gender gaps favouring boys is positively related to school-level characteri...

Selection to programmes for gifted students in Israel, performed in the second grade, relies on raw ability and achievement test scores, irrespective of age, thereby ignoring the well-known effect of within-grade age differences on test scores. Employing the entire cohort of third graders of legal age (67,366 students, 1.4% of whom were enrolled in a gifted programme) in 2011, this study examined the relationship between chronological age and the probability of acceptance to programmes for the gifted in Israel. The results reveal an almost perfect correlation between chronological age and the probability of being selected for a gifted programme (r2 = 0.92): older students have approximately 3.5 times greater chance of acceptance than younger students. Reliance on raw ability and achievement test scores, unadjusted for age, in selection to gifted students programmes is thus shown to be biased in favour of older students and to unintentionally discriminate against younger students.

Predictive validity considerations in selection dictate choice of the predictor with the highest predictive validity. Implementation of this principle in any specific selection process inevitably entails choice between imperfectly correlated alternative predictors, real or hypothetical, which are equivalent in terms of predictive validity. We show that use of different predictors necessarily results in the admission of partially overlapping groups of candidates. The amount of non-overlap is inversely related to the correlation between predictors and to the selection ratio. For typical values of these factors, the amount of non-overlap is considerable (50%-80%). Thus, the predictor’s identity affects the identity and characteristics of the admitted candidates independently of its validity and can be intentionally used to bias the selection outcomes in a desired direction.

The regression-based discrepancy definition of learning disabilities has been suggested by Rutter and Yule as an improvement of the well-known and much criticized achievement–intelligence discrepancy definition, whereby the examinee’s predicted reading attainment is substituted for the intelligence score in the discrepancy expression. Even though the regression-based discrepancy definition has been with us for more than 30 years, critical examination of this approach is scarce. This article fills this lacuna by examining the implications of two variables in the model on the diagnosis of learning disabilities: (a) the effect of predictive validity on the proportion of examinees identified as learning disabled, and (b) the effect of the predictor’s identity on the identity of the examinees diagnosed with learning disabilities. Implications of these effects concerning the validity of the regression-based discrepancy model and of the results of its implementation are discussed.

Fiedler and Kareev (2006) showed that small samples can, in principle, outperform large samples in terms of the quality of contingency-based binary choice. The 1st part of this comment critically examines these authors' claim that this small sample advantage (SSA) contradicts Bernoulli's law of large numbers and concludes that this claim is unwarranted. The 2nd part of the comment provides insight into the etiology of the SSA and points to the following as necessary conditions for the SSA's occurrence: (a) the statistical invalidity of the underlying threshold-based decision algorithm and (b) the particular payoff scheme underlying the definition of the decisions' quality. Together, these 2 factors explain how better information provided by larger samples is translated into worse decisions.

Insufficiently regressive intuitive predictions have been attributed to mistaken reliance on the representativeness heuristic. In contrast, we suggest that intuitive predictions stem from a conceptualization of ‘goodness of prediction’ that differs from the accepted statistical definition in terms of error minimization, namely, ecological validity—that is, representation of the substantive characteristics of the predicted variable Y and its distribution as well as of the relationship between Y and the predictor X—rather than minimization of prediction errors. Simultaneous satisfaction of the above representation requirements is achieved by multivalued prediction: The prediction of different Y′ values for the same X value, resulting in conditional distributions Y′|X for at least some X values. Empirical results supporting this hypothesis are presented and discussed. Copyright © 2008 John Wiley & Sons, Ltd.

Using the “between-grade levels” regression discontinuity design, this study examined the hypothesized differential sensitivity of logico-mathematical (LM) and infralogical (IL) operational tasks to the effects of chronological age and first grade schooling in a sample of 580 1st and 2nd grade Israeli children. The results indicate that the development of logico-mathematical operational skills (classification, class inclusion and transitivity) is mainly attributable to schooling. In contrast, the effect of schooling on the development of conservation of mass, liquid quantity and number (clearly an infralogical task) is negligible; acquisition of conservation is almost exclusively due to maturation and out-of-school experiences. The results support the theoretical predictions derived from French-Swiss research of the last two decades and are inconsistent with claims regarding the specificity of schooling effects to tasks that are taught in school.

This paper integrates recent meta-analytical findings regarding group differences in joband educational-related criteria and cognitive ability measures used as predictors in personnel selection and selection to higher education institutions. The findings suggest that cognitive ability measures reveal much higher group differences than the corresponding between-group differences in job- and educational-related criteria. One possible explanation for these differential gaps is that cognitive ability measures are objective and standardized while the typical measures used as job- and-educational related criteria are non-standardized subjective evaluations of job performance and academic achievement. While these findings are consistent with unbiased prediction or over-prediction for lower scoring groups, they imply that selection is biased against them. Implications and future research are discussed.

This article challenges Yaakov Kareev's (1995a, 2000) argument regarding the positive bias of intuitive correlation estimates due to working memory capacity limitations and its adaptive value. The authors show that, under narrow window theory's primacy effect assumption, there is a considerable between-individual variability of the effects of capacity limitations on the intuitive assessment of correlation, in terms of both sign and magnitude: Limited capacity acts as an amplifier for some individuals and as a silencer for others. Furthermore, the average amount of attenuation exceeds the average amount of amplification, and the more so, the smaller the capacity. Implications regarding the applicability and contribution of the bias notion in this context and the evaluation of the adaptive value of capacity limitations are discussed.

Despite its intuitive appeal and popularity, Thorndike’s constant ratio (CR) model for unbiased selection is inherently inconsistent in n-free selection. Satisfaction of the condition for unbiased selection, when formulated in terms of success/acceptance probabilities, usually precludes satisfaction by the converse probabilities of rejection/failure, and vice versa. This paper suggests that this inconsistency is unavoidable due to the (negative) linear relation between “percent accepted” (P) and “percent rejected” (1 –P ), which does not preserve ratios and, thus, invalidates the conceptualization and measurement of selection bias in ratio terms. Therefore, we propose to substitute the CR model with a constant difference (CD) model for the definition and measurement of selection bias, and show the latter’s underlying rationale and its applicability in both n-free and fixed-n selection. The psychometric literature dealing with bias in selection has focused mainly on personnel and educational selection where the criterion of merit Y—a measure

This study estimated the independent effects of age and schooling in grades 7–9 on scores obtained on invalid conditional and class syllogisms. The results, which point to a negative, albeit small, effect of out-of-school experience and to a sizeable positive effect of schooling, replicate previous findings in a different age range and support the counterintuitive hypothesis that accumulated daily experience with conditionals has a negative effect on the development of conditional reasoning, and that improved performance on invalid problems with age is entirely attributable to schooling. Contrary to most cognitive tasks, therefore, in which schooling operates in the same direction as out-of-school experience, in this case, schooling breaks daily-life interpretational habits, and therefore, is critical for development.

How does schooling affect the development of intelligence in children? How should the amount of schooling be considered when developing norms for turning intelligence test performance into IQ scores?

The Kaufman Ability Battery for Children (K-ABC) is predicated on the distinction between intelligence (measured by the Mental Processing scale) and achievement (measured by the Achievement scale). This article questions the validity of this distinction and the interpretation of the K-ABC’s Mental Processing score as a measure of “pure” intelligence, uncontaminated by schooling. Schooling was found to affect raw scores on all Mental Processing subtests in the Israeli standardization sample of the K-ABC. More-over, for five of the eight subtests, the effect of schooling exceeded the effect of age. No difference was found between the Mental Processing and Achievement scales in the relative magnitude of the age and schooling effects.

This study investigates the relationship between age at immigration and educational achievement at age 14 among all the students (about 45,000) who immigrated to Israel between 1952 and 1970. The relationship is examined for verbal and nonverbal components of an achievement test, in subpopulations defined by gender and ethnic background (Westernversus Eastern-born children). The findings indicate a monotonic decrease in achievement as a function of immigration age starting at the age of 7. This decrease is considerably stronger for the Verbal subtest than for the Mathematical subtest, particularly in the Western group. These results refute the vulnerable age hypothesis: They support the view that the foreign language acquisition factor plays a central role in the relationship between age of immigration and scholastic achievement, and are consistent with the expectation of a monotonic decline in achievement as age of immigration increases (and length of residence decreases). Hence, when immigration involves the need to learn a new language, the drop in school achievement is likely to be particularly marked in subject areas requiring higher levels of mastery of the language of instruction. Finally, the results suggest that age 7 may represent a critical age for the scholastic achievement of immigrant students.

Empirical examination of the relation between prediction bias and selection bias has been hampered by the lack of a valid definition of selection bias. In this article, we show that such a definition is possible in the special case of fixed-n selection, where the only reason for rejecting otherwise acceptable applicants is that their number exceeds the number of available places. Using the proposed definition, the empirical relation between prediction bias and selection bias with respect to prominent social groups is examined. The results indicate that although the 2 biases are related, the relation is not isomorphic: First, it is mediated by the selection ratio; second, for most selection ratios, its strength is only moderate. Thus, both the direction and the magnitude of prediction bias may be misleading as a basis for determining the corresponding direction and magnitude of selection bias. In particular, lack of prediction bias does not entail lack of selection bias. Hence, examination of selection bia...

This paper introduces a way of understanding correlation coefficients based on the notion of scaled-score similarity (i.e., closeness to identity). It is suggested that the proposed framework has intuitive appeal. It points to the equivalence among various correlation coefficients by showing that correlation problems involve the notion of agreement between interval or ordinal measures expressed on different scales. This paper introduces a common conceptual framework for the definition and interpretation of correlation coefficients in terms of degree of similarity (i.e., closeness to identity) between scaled scores (e.g., ranks, percentiles, or z-scores). It is suggested that this framework facilitates the teaching and understanding of the correlation concept by pointing to the conceptual equivalence among various correlation coefficients. The model is illustrated in the specific case of ranks, z-scores and percentiles.

In order to prevent misleading conclusions based on spurious observed effects—especially seductively large ones—Robinson and Levin (1997) suggested a two step approach to the reporting and evaluation of empirical results. According to the two-step model, the evaluation of the magnitude and substantive significance of obtained effects should be conditional upon their statistical significance: Author should first indicate whether the observed effect is statistically improbable, and only if it is should they then indicate how large or important it is. The purpose of their paper is (a) to show that even through the underlying intention—to prevent unwarranted evaluation of spurious observed effects—is a laudable one, the two-step approach is inappropriate for this purpose, and (b) to reiterate the preferred approach, namely increased sample size and computation of confidence intervals.

The study investigated gender differences among 11,000 Israeli children in Grades 4–6 with respect to verbal, spatial, and mathematical ability, as measured by 12 intelligence tests. Consistent differences in score variance were found across grades for 11 of the 12 tests. In each of these tests the variance for boys exceeded that for girls by 10%–20%. With respect to mean achievement, consistent cross-grade differences were found only for mathematical ability, where boys had the edge (about 0.20 SD). These findings diverge from those of recent American studies, which found no gender differences in any of these realms. Furthermore, they differ from the results of earlier Israeli studies in that the gender gap is limited to mathematical ability, and its size is much smaller. The revealed gender gap can be partially attributed to differences in response strategy: girls were found to be more likely to skip items for which they lack an answer (i.e., to take fewer risks in guessing). This implies that the performance of girls on intelligence tests will improve if they are encouraged to dare to guess.

Because it is impossible to experiment with school attendance, the erect of formal education as opposed to chronological age on the development of transitive inference has never been investigated empirically. A recent quasi-experimental paradigm, which allows for disentangling the net erects of age and schooling, can help overcome this difficulty. The paradigm is applied to the estimation of the independent erects of age and schooling in Grades 5 and 6 on raw scores obtained in a 3-term series test. Results point to schooling as a major factor underlying the increase of scores as a function of age

In light of the outdatedness of empirical research on IQ constancy among gifted children, and with the aim of examining possible cross cultural differences, the present study investigated the issue within the Israeli context. Specifically, we analyzed the constancy of IQ scores on the WISC‐R test for 161 kindergarteners through fourth graders identified as gifted by the Jerusalem Psychological Service in 1981/82‐ 1983/84. Assessment of IQ constancy was based on a retest administered to subjects 1–4 years after the first test. Results showed that 86% of the children in the sample were defined as gifted also on retest. Mean absolute differences between testings ranged from 1/3 to 1/2 SD (5–8 IQ points) for Verbal, Performance and Full‐scale IQ scores, and from 1/2 to 3/4 SD for subtest scores. On the whole, Performance scores remained constant, while Verbal scores tended to decline. There were no consistent differences attributable to age of identification or measurement interval.

The independent influences of aging and schooling on the development of phonological awareness were assessed using a between-grades quasiexperimental design. Both schooling (first grade) and aging (5–7 years) significantly improved children's performance on tests of phonemic segmentation, but the schooling effect was four times larger than the aging effect. The schooling effect was attributed to formal reading instruction, whereas the aging effect probably reflects natural maturation and informal exposure to written language. These data support a strong mutual relation between reading acquisition and phonological awareness.

The probabilities of the two types of classification errors in competency testing are not equally manipulable. Whereas testers can successfully minimize the probability of Type II error (misidentification of true "nonmasters" as "masters"), they are much less able to do so relative to Type I error (misidentification of true "masters" as "nonmasters"). Consequently, the proportion of identified nonmasters is likely to be artifactually high. The currently used method for coping with this problem is the arbitrary setting of observed-score standards below 100%. This paper offers an alternative solution which does not involve arbitrary decisions -namely, self-correction of wrong answers. The paper presents the rationale underlying this solution and discusses its application.

This paper considers the efficiency of psychological assessment with regard to transition from kindergarten to first grade by examining the relation between information available to the psychologist mainly, the kindergarten teacher's recommendation and the results of tests administered during the assessment procedure and the ultimate recommendation reached by the psychologist (retention in kindergarten versus entrance into school). Employing a sample of Israeli kindergarten children referred to the psychological-educational service, the study found no relation between the psychologist's recommendation and test scores and a strong correspondence between it and the kindergarten teacher's recommendation. Agreement was particularly high regarding retention in kindergarten (the vast majority of recommendations) and with respect to the younger children in the cohort (July-December), who constituted 80 percent of the referral population. From an organizational standpoint, these findings suggest that psychological assessment is superfluous, at least with respect to the vast majority of referrals.

Narrow Window theory, suggested by Y. Kareev ten years ago, has so far focused on one central implication of the limited capacity of working memory on intuitive correlation estimation, namely, overestimation of the distal population correlation. This paper points to additional and perhaps more dramatic implications due to the large dispersion of intuitive estimates: (a) large estimation errors, possibly causing overestimation of negligible ρ s, misses of strong ρ s, and distorted hierarchies of the ρ s between different pairs of variables; and (b) large interpersonal differences in the estimation of any given ρ and highly incongruent hierarchies of estimated correlations between different pairs of variables. These implications impede both individuals’ adaptation to the empirical world and communication among themselves.

The aim of this study was to investigate whether there are increased levels of the inflammatory cytokines IL-1beta, IL-8, and TNFalpha in the gingival crevicular fluid (GCF) of erupting primary teeth. This increase could explain such clinical manifestations as fever, diarrhea, increased crying, and sleeping and eating disturbances that occur at this time.Sixteen healthy children aged 5 to 14 months (mean=9.8 months) were examined twice a week over 5 months. Gingival crevicular fluid samples were taken from erupting teeth. As a control, GCF was collected from the same teeth 1 month later. Cytokine production was measured by ELISA. Signs and clinical symptoms were listed. Pearson correlation coefficients were used in the comparisons described below. A paired t test was used to analyze the same variable at different times.Fifty teeth of the 16 children were studied. GCF samples were collected from 21 of these teeth. Statistically significant differences (P.05) were found with regard to the occurrence of fever, behavioral problems, and coughing during the teething period and the control period. During the control period, 72% of the children did not exhibit any clinical manifestations, whereas during the teething period only 22% of the children did not exhibit any clinical manifestations. The study revealed high levels of inflammatory cytokines during the teething period, with a statistically significant difference in TNFalpha levels (P.05) between the teething period and the control period. Correlations were found between cytokine levels and some of the clinical symptoms of teething: IL-1beta and TNFalpha were correlated with fever and sleep disturbances; IL-beta and IL-8 were correlated with gastrointestinal disturbances; IL-1beta was correlated with appetite disturbances.Cytonkines appear in the GCF of erupting prmary teeth. The cytokine levels are correlated to some symptoms of teething.

In order to prevent misleading conclusions based on spurious observed effects - especially seductively large ones - Robinson and Levin (1997) suggested a two-step approach to the reporting and evaluation of empirical results. According to the two-step model, the evaluation of the magnitude and substantive significance of obtained effects should be conditional upon their statistical significance: Authors should first indicate whether the observed effect is statistically improbable, and only if it is should they then indicate how large or important it is. The purpose of this paper is (a) to show that even though the underlying intention - to prevent unwarranted evaluation of spurious observed effects - is a laudable one, the two-step approach is inappropriate for this purpose, and (b) to reiterate the preferred approach, namely increased sample size and computation of confidence intervals.

Two criteria for the evaluation of intra-individual subtest score differences have been suggested in the literature, namely "statistical significance" and "abnormality." The procedures for determining them involve a z-score transformation of the observed difference d—the division of d by the standard error of measurement of the difference (uMd) in the case of statistical significance, and by the standard deviation Cd of the observed differences in the case of abnormality. As oMd

The universal nature of school attendance precludes the experimental investigation of the absolute effect of schooling on achievement. The available empirical evidence consists of the results of post hoc analyses of the within-grade-level relationship of students’ achievement to the quality and quantity of schooling. This paper suggests a between-grade-levels quasi-experimental approach to the investigation of school’s effect on achievement, involving the estimation of this effect by means of a regression discontinuity design. The paper presents the rationale underlying the suggested approach and demonstrates its application to the estimation of the effect of schooling at grade 2 on achievement in mathematics and reading comprehension.

The las t twenty years have witnessed an ever-increasing awareness of the "unit of analysis" problem and a growing acceptance of the school as the appropriate object of research. Among the var ious school characterist ics , achievement is the most prominent. Because instruct ion is usual ly organized according to grade level and subject mat ter , the m e a s u r e m e n t of school achievement involves the adminis t ra t ion of a different achievement test for each combinat ion of grade and subject matter, with the resul ts for each test providing school mean scores. Therefore, the conceptual iza t ion of school a c h i e v e m e n t n e c e s s a r i l y i nvo lves t he s e t or prof i le of the grade-by-subject-matter school means.

The purpose of this paper is to point out the logical, theoretical and statistical problems associated with the computation of the "pooled-within" correlation coefficient in partially segregated educational systems. It is argued that, since a between-individual-within-classes relationship of Y to X cannot possibly exist in classes which are entirely homogeneous with respect to either X or Y, the question concerning this relationship should be addressed only to the subsample of heterogeneous classes. Consequently, the measurement of this relationship should be restricted to this subsample, thus yielding a new coefficient, rwH, which is different from the usual coefficient, both conceptually and numerically. The remainder of the paper deals with the theoretical and statistical problems associated with the computation of rwH instead of rw, its interpretation as a sample statistic and its functional relationship to rB and rT, the "between groups" and "total" correlation coefficients, respectively.

Though the testing for the statistical significance of subtest score differences yields a considerable proportion of nonsignificant results (e.g., about two-thirds of the V-P IQ differences on the WISC-R, assuming α = .05), the interpretation of such results has been neglected. This paper examines the implications of the decision rule as they concern statistically nonsignificant results. The acceptance of the null hypothesis after a nonsignificant difference is shown to result in a noticeable proportion of both Type II and classification errors. Thus, in view of its low statistical power and the considerable proportion of classification errors involved, the widespread use of significance testing of subtest score differences is questioned; it cannot be justified in terms of a very extreme loss function, which emphasizes the minimization of the α-probability of Type I errors at the expense of the β-probability of Type II errors. Two other explanations for the universal acceptance of the significance testing of subtest score differences are suggested and critically examined.

Significance testing is widely used as one way to evaluate subtest score differences. Nevertheless, misinterpretations of statistically significant differences are frequent, due to basic misunderstandings of the underlying model. This article reviews the significance testing model and discusses the interpretations and misinterpretations of statistically significant differences.