Abstract

This study investigated the convergent and discriminant validity of the naming subtest of the Neuropsychological Assessment Battery (NAB), a measure of word-finding. Seventy community-dwelling adults age 60 and above completed the NAB naming test (Forms 1 and 2), the Boston Naming Test (BNT), and other measures of verbal and visual memory, visuoperceptual skills, processing speed, and abstraction. The NAB naming test correlated highly with the BNT and with established measures of memory. The BNT correlated more strongly with education and with sex. The BNT correlated more strongly with a measure of visuoperceptual skills than the NAB naming test did, suggesting that visuoperception is more involved in the BNT. Divergent validity of the NAB naming was demonstrated by a lack of correlations with less-related measures. Findings suggest that the NAB naming test possesses convergent and divergent validity as a measure of word-finding.

Introduction

A commonly found symptom of Alzheimer's disease is anomia, or difficulty generating the names of objects or retrieving words (Hodges, Patterson, Graham, & Dawson, 1996; Nebes, 1989; Salmon & Bondi, 2009; Thompson-Schill, Gabrieli, & Fleischman, 1999; Williams, Mack, & Henderson, 1989). It is thus paramount for neuropsychologists working with older adults to evaluate word-finding ability. This is typically completed through confrontation naming tests, in which the patient is presented with a picture of an object, which she or he must name. Measures of word-finding are one of the most commonly used measures utilized by neuropsychologists (Rabin, Barr, & Burton, 2005).

The most commonly used measure of word-finding is the Boston Naming Test (BNT) (Rabin et al., 2005). This test was originally created in 1978 (Kaplan, Goodglass, & Weintraub, 1978) and has stayed largely true to its format since that time (Kaplan, Goodglass, & Weintraub, 2001). Patients with Alzheimer's disease show impairment on it (Henry, Crawford, & Phillips, 2004; Testa et al., 2004). Scores on the test can be used to distinguish between dementia due to clinically diagnosed Lewy body disease and Alzheimer's disease (Williams et al., 2007). It has been shown to have 68% sensitivity and 70% specificity in the identification of cognitive impairment in older adults (Lichtenberg, Vangel, Kimbarow, & Ross, 1996). However, like many neuropsychological measures, the BNT suffers from weaknesses that were not well known when the test was first created. It is highly correlated with education (Hawkins & Bender, 2002; Ross & Lichtenberg, 1998; Zec, Burkett, Markwell, & Larsen, 2007) and affected by cultural assimilation (Lucas et al., 2005; Manly et al., 1998; Ross & Lichtenberg, 1998), which can limit its validity as a measure of word-finding ability. Part of the reason for this is that some of the items (e.g., a protractor, a compass, a pelican) may have never been seen by examinees with lower levels of education or with less assimilation into American culture. The demonstrated effects of education and cultural assimilation have led to the need for multiple normative databases to be established in order to prevent misdiagnosis of older patients as impaired. The Mayo Older African American Normative Studies (MOAANS; Lucas et al., 2005) provided valuable normative data exclusively for African Americans. Using these norms, a clinician can find that a 75-year-old African American scoring at the 50th percentile would have been classified as scoring at the 16th percentile using the previously published Mayo Older American Normative Studies (MOANS; Ivnik, Malec, Smith, Tangalos, & Peterson, 1996). Some of the objects pictured on the test (e.g., a noose) would probably not be included on culturally sensitive measures created today. Another administration difficulty for the test is that the order of items has been shown to be weakly related to their difficulty level (Lichtenberg, 1998), leading to uncomfortable test situations in which clinicians must proceed with test administration despite repeated failures by the examinee that do not meet the discontinue criteria. Lastly, the stimuli consist of line drawings that can sometimes be misperceived by patients, especially those with poor vision. For example, the drawing of a pretzel is often misperceived as a snake. Indeed, Williams and colleagues (2007) demonstrated the occurrence of misperception of test stimuli when they found that patients with clinically diagnosed dementia with Lewy bodies made more visuoperceptual errors on the test than patients with clinically diagnosed Alzheimer's disease.

The BNT can be a lengthy measure to administer, taking up to 20 min to administer to older adults (Strauss, Sherman, & Spreen, 2006). This problem has been addressed by the creation of short forms of the test consisting of 15 items (Fastenau, Denburg, & Mauer, 1998; Graves, Bezeau, Fogarty, & Blair, 2004; Lansing, Ivnik, Cullum, & Randolph, 1999; Mack, Freed, Williams, & Henerson, 1992; Saxton et al., 2000) or 30 items (Williams et al., 1989). The second edition of the test (Kaplan et al., 2001) includes Mack et al.'s short form option.

The neuropsychological assessment battery (NAB; Stern & White, 2003) has five modules designed to assess attention, visuospatial skills, language, memory, and executive functioning. Each module has two forms (Forms 1 and 2) to enable repeat testing of a patient. The language module contains measures of oral production, auditory comprehension, naming, word and sentence comprehension, and writing. The NAB has both demographically corrected norms (N = 1448) and a subsample of norms (N = 950) that matches the U.S. population on the variables of age, education, gender, ethnicity, and geographic region. Of the 1448 participants, 711 completed Form 1 and 737 completed Form 2. The naming subtest of the language module bears the most similarity to the BNT. It involves similar confrontation naming procedures of requesting the examinee to generate the names of objects on their own, and then providing a semantic cue (e.g., “It's a type of animal”) if they cannot generate the name, and then providing a phonemic cue (i.e., the first sound of the word) if the examinee still cannot generate the name. There are several notable differences between the NAB and the BNT. The NAB naming test involves showing the examinee color photographs of objects, as opposed to black and white line drawings. The NAB naming test has fewer items (31 as opposed to 60), and all items are administered to test-takers, with no setting of a basal level or discontinue criteria. Patients are allowed 10 s to freely recall the name of the item, and then are given 5 s after the semantic cue and 5 s after the phonemic cue, for a total of 20 possible seconds per item. The BNT, on the other hand, according to commonly accepted instructions (Strauss et al., 2006) provides 20 s for the examinee to freely recall the name, followed by 20 s after the semantic cue and 20 more after the phonemic cue. This can lead to a lengthy administration (up to 1 min per item), leading some examiners to break standardization to speed up the testing process.

The NAB naming test holds promise as a measure of word-finding ability, and some validity information for the test battery is available in its technical manual (White & Stern, 2003). The manual notes that age, education, and sex account for 3.4%, 1.1%, and 0.2% of the variance, respectively, in raw NAB naming scores, in a multiple regression analysis. When the raw scores are normed, the variance due to these variables is removed. White and Stern report α reliability coefficients of .79 and .73 for Forms 1 and 2, respectively. Alternate forms reliability of the two forms was tested with generalizability theory and found to be .72. The manual reports that the NAB naming test correlated .56 with the BNT in a non-impaired sample, and .76 in a sample with aphasia. White and Stern note that these analyses are only the starting point in the validation of these tests, and note that future research should extend these findings.

White and Stern (2003), in their technical manual, did not report several analyses that the current study sought to conduct. First, although correlations between the NAB naming test and other index scores from the NAB, and the Reynolds Intelligence Screening Test (Reynolds & Kamphaus, 2003) are presented, correlations with measures of other relevant domains (e.g., memory) from other test batteries are not presented. Second, evidence of discriminant validity is not presented. For example, this could be established by correlating the NAB naming test with measures of domains that should be unrelated, or at least not highly related (e.g., processing speed, visuospatial skills, or abstraction). Third, the current study aims to conduct these analyses not only with the NAB naming test, but also with the BNT, in order to assess how the convergent and discriminant validity of the NAB naming test compares with that of the BNT. The current study also aimed to explore the relationship between the NAB naming test and demographic variables of age, education, and gender and compare these relationships to those with the BNT. White and Stern do not report how much of the variance in BNT performance was explained by these demographic variables in their normative sample.

The purpose of this study was to independently evaluate the convergent and discriminant validity of the NAB naming test as a measure of word-finding in a sample of older adults. It was hypothesized that both forms of the NAB naming test would correlate highly with the BNT, establishing convergent validity. The second hypothesis was that the NAB naming test would correlate moderately with measures of memory, also helping to establish convergent validity. This might be expected given the proximity of temporal lobe structures responsible for memory and language (Blumenfeld, 2002). Third, the NAB naming test was predicted to be correlated with a measure of visuoperceptual skills, due to the visual nature of the task, but less so than the BNT, due to its stimuli consisting of line drawings as opposed to color photographs. Lastly, it was hypothesized that the NAB naming test would correlate weakly, if at all, with measures of less-related abilities of processing speed and executive functioning and with the level of education and gender. Weak correlations with these less-related variables would help to establish divergent, or discriminant, validity.

Materials and Methods

Participants

Demographic characteristics of the sample are listed in Table 1. Study participants included 70 adults, age 60 and over, who were recruited from a registry of community-dwelling older adults who expressed interest in participating in research. All participants lived in an urban area in the western USA and were paid $20 for approximately 90 min of their time. Their self-described ethnicity was 96% European American and 4% Hispanic American, and 59% were women. They were generally physically healthy, as indicated by a mean of 1.2 self-reported medical problems (SD = 1.5; range of 0–7) on the comorbidity index (Charlson, Pompei, Ales, & MacKenzie, 1987). The most common medical problems reported were diabetes (N = 17), connective tissue disease (e.g., arthritis; N = 12), chronic obstructive pulmonary disease (N = 9), cancer (N = 8), history of myocardial infarction (N = 7), cerebrovascular disease (N = 7), ulcer disease (N = 4), and peripheral vascular disease (N = 4). All participants provided informed consent, and this study was approved by the university's institutional review board. English was the primary language of all participants.

Table 1.

Demographics and mean scores on neuropsychological measures

Variable M SD Range 
Age 75.4 7.4 60–88 
Years of education 15.1 3.5 8–20 
Comorbidity index 1.20 1.5 0–7 
NAB naming form 1 30.0 1.4 26–31 
NAB naming form 1 T-score 55.6 7.0 35–64 
NAB naming form 2 29.9 1.3 25–31 
NAB naming form 2 T-score 55.1 7.8 27–64 
BNT 56.1 3.8 44–60 
BNT scaled score 13.0 2.7 8–17 
Judgment of line orientation 23.8 4.0 13–30 
Judgment of line orientation scaled score 12.0 2.7 6–17 
BVMT-R total recall 16.4 5.6 7–27 
BVMT-R total recall T-score 42.4 9.6 26–61 
CVLT-2 Trials 1–5 43.7 11.2 21–70 
CVLT-2 Trials 1–5 T-score 54.1 10.2 34–77 
D-KEFS 20 questions abstraction 22.9 11.9 0–58 
D-KEFS 20 questions abstraction scaled score 9.8 2.8 3–18 
WAIS-3 processing speed index score 108.3 13.1 81–137 
Variable M SD Range 
Age 75.4 7.4 60–88 
Years of education 15.1 3.5 8–20 
Comorbidity index 1.20 1.5 0–7 
NAB naming form 1 30.0 1.4 26–31 
NAB naming form 1 T-score 55.6 7.0 35–64 
NAB naming form 2 29.9 1.3 25–31 
NAB naming form 2 T-score 55.1 7.8 27–64 
BNT 56.1 3.8 44–60 
BNT scaled score 13.0 2.7 8–17 
Judgment of line orientation 23.8 4.0 13–30 
Judgment of line orientation scaled score 12.0 2.7 6–17 
BVMT-R total recall 16.4 5.6 7–27 
BVMT-R total recall T-score 42.4 9.6 26–61 
CVLT-2 Trials 1–5 43.7 11.2 21–70 
CVLT-2 Trials 1–5 T-score 54.1 10.2 34–77 
D-KEFS 20 questions abstraction 22.9 11.9 0–58 
D-KEFS 20 questions abstraction scaled score 9.8 2.8 3–18 
WAIS-3 processing speed index score 108.3 13.1 81–137 

Notes: M = mean; SD = standard deviation; NAB = Neuropsychological Assessment Battery; BNT = Boston Naming Test; BVMT-R = Brief Visuospatial Memory Test, revised; CVLT-2 = California Verbal Learning Test, second edition; D-KEFS = Delis–Kaplan Executive Function System; WAIS-3 = Wechsler Adult Intelligence Scale, third edition.

Upon reviewing the distribution of test scores in the sample, it is possible that the sample included several people with dementia and/or mild cognitive impairment. Indeed, two participants self-reported having dementia on the comorbidity index, though the accuracy of self-report of this particular diagnosis can be questioned. In any event, it was considered valuable to include participants with a wide range of cognitive functioning for this study. Having a wide range of cognition would make the sample more representative of the population over 60 and would increase the variance under study.

Materials

The NAB naming test, Forms 1 and 2, and the 60-item BNT were administered. On the BNT, the total score was the total spontaneously correct plus those correct after the stimulus cue. No credit was given toward the total score for correct items after the phonemic cue.

In addition, participants completed the following measures. Raw and normed scores for the above-mentioned measures are presented in Table 1. Normed scores were either obtained from the test manuals (for the NAB naming, CVLT-2, BVMT-R, WAIS-3 PSI score, and D-KEFS 20 questions) or from the MOANS (Ivnik et al., 1996) for the BNT and judgment of line orientation. Data analyses were conducted with raw scores.

  • The California Verbal Learning Test, Second Edition. The California verbal learning test, second edition (CVLT-2; Delis, Kramer, Kaplan, & Ober, 2000), is a measure that assesses both verbal learning and memory. The CVLT-2 is the fourth most commonly used measure by neuropsychologists, and the second most commonly used measure of memory, after the Wechsler Memory Scale (Rabin et al., 2005). Participants are read a 16-word list five times, and after each trial asked to freely recall as many of the words as possible. This is followed by other components of the test. This study utilized the total number of words recalled in Trials 1–5 as a measure of verbal memory, for the following reasons. This measure has high internal consistency based on split-half reliability, r = .94. This measure also has demonstrated validity as a measure of verbal memory (Delis et al. 2000). The Trials 1–5 total score is the most normally distributed measure on the CVLT-2. It is for this reason that raw scores on Trials 1–5 total are converted into T-scores on the CVLT-2, whereas raw scores on the other CVLT-2 measures (e.g., delayed recall) are converted into z-scores because they are less normally distributed (Delis et al. 2000).

  • Brief Visuospatial Memory Test–Revised. The brief visuospatial memory test, revised (BVMT-R; Benedict, 1997), is a measure of visual learning and memory. Participants are asked to study a sheet with six figures on it for 10 s, and then draw from memory all the figures accurately and in their correct location. This is repeated for a total of three trials. Following a 25 min delay, participants are asked to freely recall and again draw the six figures. This measure has high test–retest reliability (r = .80) for the total recall score for Trials 1–3, and the total recall score may be the most sensitive measure of visual memory on the BVMT-R (Benedict, 1997). The BVMT-R total recall score has strong criterion validity, with high correlations with the Rey Complex Figure Test recall trial and with the Hopkins Verbal Learning Test (Benedict, Schretlen, Groninger, Dobraski, & Shpritz, 1996).

  • Judgment of Line Orientation. The Judgment of Line Orientation Test (Benton, Sivan, Hamsher, Varney, & Spreen, 1994; Benton, Varney, & Hamsher, 1978) is a measure of visuoperceptual functioning. It consists of 30 stimuli with one page with two lines and a second page with 11 lines. Participants are asked to compare the two pages and report which two lines on the second page point in the same direction and are in the same location as the two lines on the first page. This measure has high split-half reliability for adults (r = .84–.91; Strauss et al., 2006) and demonstrated validity (Riccio & Hynd, 1992).

  • WAIS-3 Processing Speed Index. The processing speed index (PSI) of the WAIS-3 (The Psychological Corporation, 1997) is comprised the digit symbol and symbol search subtests. Raw scores from both subtests are first converted into scaled scores using tables from the WAIS-3 Administration and Scoring Manual. Scaled scores are then summed and transformed into index scores through tables in the manual. The split-half reliability coefficient for the PSI is high (.88), based on the standardization sample. PSI scores are also moderately correlated with Working Memory Index scores (r = .51).

  • Delis–Kaplan Executive Function System (D-KEFS) 20 questions subtest. The D-KEFS 20 questions subtest (Delis, Kaplan, & Kramer, 2001a) is a measure of executive functioning or abstraction. In this measure, participants are given a sheet that has 30 multicolored items on it and are to ask the fewest number of yes/no questions possible to figure out which item the examiner has picked. The 30 items can be placed into categories (e.g., animals, forms of transportation), which is thought to involve abstraction skills. Placing the items into categories enables the participant to guess the item utilizing fewer questions. Participants are asked to identify four items total for a total of four trials. The initial abstraction score for this measure has high split half reliability, r = .72–.87 (Delis, Kaplan, & Kramer, 2001b). This score was significantly lower among patients with frontal lobe lesions than in age- and education-matched controls (Baldo, Delis, Wilkins, & Shimamura, 2004).

Several other measures were included in the test battery as parts of other research studies, but were not included in the current study. Other executive function measures (e.g., the trail making, verbal fluency, and tower tests from the D-KEFS) were not included because they each are timed tests. Thus, any correlation between these measures and the naming tests could have been attributed to shared variance in processing speed. The 20 questions subtest is not timed and was thus considered to be even less related to naming and was included to demonstrate divergent validity.

Procedures

Participants were individually administered the neuropsychological test battery as part of a larger study, in a neuropsychological research lab. The three naming tests were administered in a random order for each participant to control for order or practice effects.

Statistical Analyses

Data analyses consisted primarily of Spearman's ρ and Pearson product moment correlations calculated among raw scores from the neuropsychological measures (with the exception of the WAIS-3 PSI score, which is not a raw score). The NAB and BNT were also correlated with demographic variables. In addition, Steiger's Z (two-tailed) was used to compare the strengths of correlations between the naming tests and other variables.

Results

Mean scores on the NAB naming test, Forms 1 and 2, as well as the BNT, are included in Table 1. Scores on all three measures were negatively skewed, which likely attenuated the size of the correlations and necessitated the calculation of Spearman's ρ rather than Pearson correlations with the naming tests. The median scores were 57 on the BNT, 31 on NAB naming Form 1, and 30 on NAB naming Form 2. The BNT displayed a larger range of scores than the NAB naming tests. There are several overlapping items on the BNT and NAB naming tests; this may have slightly decreased the range of scores.

Spearman's ρ correlations were calculated between demographic variables and the NAB and BNT. Age correlated with NAB naming Form 1, r = −.32, p < .01, and Form 2, r = −.28, p < .05, but not with the BNT, r = −.17, p = .16. Years of education correlated with the BNT, r = .40, p < .01, followed by NAB naming Form 1, r = .32, p < .01, and Form 2, r = .28, p < .05. Steiger's Z was used to calculate whether the Spearman's ρ correlations between the naming tests and education were significantly different. The correlation with education was not significantly different between the BNT and NAB naming Form 1, Z = 0.73, p > .05, or between the BNT and NAB naming Form 2, Z = 1.12, p > .05. Sex correlated with the BNT, r = .31, p < .01, such that men obtained higher scores, but sex did not correlate with the NAB naming test (both ps > .58).

Forms 1 and 2 of the NAB naming test correlated with each other, r = .50, p < .001. All other correlations are presented in Table 2. Forms 1 and 2 were highly correlated with the BNT, r = .45, p < .001 (Form 1) and r = .50, p < .001 (Form 2), providing evidence of convergent validity.

Table 2.

Correlations among neuropsychological measures and demographic variables

 Age Sex Education NAB form 1 NAB form 2 BNT JLO BVMT-R total recall CVLT-2 trials 1–5 D-KEFS 20 questions abstraction WAIS-3 processing speed 
Age –           
Sex .12 –          
Education −.24* .25* –         
NAB form 1 −.32** .06 .32** –        
NAB form 2 −.28* .03 .28* .50** –       
BNT −.17 .31** .40** .45** .50** –      
JLO −.25* .21 .40** .35** .20 .49** –     
BVMT-R total recall −.33** −.16 .26* .35** .35** .30* .41** –    
CVLT-2 Trials 1–5 −.46** −.44** .31** .23 .35** .26* .21 .49** –   
D-KEFS 20 questions abstraction −.18 .04 .10 .18 .21 .29* .38** .27* .23 –  
WAIS-3 processing speed .08 .05 .03 .06 .13 .34** .35** .33** .16 .36** – 
 Age Sex Education NAB form 1 NAB form 2 BNT JLO BVMT-R total recall CVLT-2 trials 1–5 D-KEFS 20 questions abstraction WAIS-3 processing speed 
Age –           
Sex .12 –          
Education −.24* .25* –         
NAB form 1 −.32** .06 .32** –        
NAB form 2 −.28* .03 .28* .50** –       
BNT −.17 .31** .40** .45** .50** –      
JLO −.25* .21 .40** .35** .20 .49** –     
BVMT-R total recall −.33** −.16 .26* .35** .35** .30* .41** –    
CVLT-2 Trials 1–5 −.46** −.44** .31** .23 .35** .26* .21 .49** –   
D-KEFS 20 questions abstraction −.18 .04 .10 .18 .21 .29* .38** .27* .23 –  
WAIS-3 processing speed .08 .05 .03 .06 .13 .34** .35** .33** .16 .36** – 

Notes: N = 70; Education = years of education; NAB = Neuropsychological Assessment Battery; BNT = Boston Naming Test; JLO = Judgment of Line Orientation; BVMT-R = Brief Visuospatial Memory Test, revised; CVLT-2 = California Verbal Learning Test, second edition; D-KEFS = Delis–Kaplan Executive Function System; WAIS-3 = Wechsler Adult Intelligence Scale, third edition. Correlations within box corresponding to naming tests are Spearman's ρ coefficients. All other correlations are Pearson's correlation coefficients.

*p < .05.

**p < .01.

Correlations were also calculated between the NAB naming tests and measures of visual and verbal memory. These abilities were considered to be distinct from, but related to, confrontation naming. The BVMT-R Trials 1–3 correlated with NAB naming Form 1, r = .35, p < .01, and Form 2, r = .35, p < .01, as well as the BNT, r = .30, p < .05. The CVLT-2 Trials 1–5 also correlated with Form 2, r = .35, p < .01, and the BNT, r = .26, p < .05. The correlation with Form 1 approached significance, r = .23, p = .06.

Judgment of line orientation correlated with Form 1, r = .35, p < .01, but not with Form 2, r = .20, p = .10. The BNT correlated more highly with judgment of line orientation, r = .49, p < .001. The judgment of line orientation correlated significantly more strongly with the BNT than it did with NAB naming Form 2, Z = 2.57, p < .05. This suggests that visuoperceptual skills are more involved with completion of the BNT than with Form 2 of the NAB naming test.

The last measures to be correlated with NAB naming were measures of processing speed and abstraction, because these were considered to be more weakly associated with confrontation naming. Weak correlations would demonstrate divergent validity. The WAIS-3 PSI score did not significantly correlate with NAB naming Form 1, r = .06, p = .63, or with NAB naming Form 2, r = .13, p = .30, but was significantly correlated with the BNT, r = .34, p < .01. The WAIS-3 PSI score correlated significantly more strongly with the BNT than the NAB naming Form 1, Z = 2.25, p < .05, but the difference in strength of correlations of the BNT and NAB naming Form 2 with the WAIS-3 PSI score was not significant, Z = 1.79, p > .05. The D-KEFS 20 questions initial abstraction score did not correlate with NAB naming Form 1, r = .18, p = .15, or Form 2, r = .22, p = .08, but did correlate with the BNT, r = .29, p < .05. The difference in strength of these correlations was not significant, p > .05.

The number of participants scoring in the impaired range was also examined. No participant scored in the impaired range on the BNT; the lowest scaled score was 8, corresponding with the 25th percentile. One participant scored more than 1 SD below the mean on NAB naming Form 1; her T-score was 35. On NAB naming Form 2, four participants scored more than 1 SD below the mean, with T-scores of 27, 33, 35, and 38. Characteristics of these four participants are presented in Table 3. Three of these participants were above the mean age of the sample, and all four had less than the mean 15.1 years of education for the sample. It is striking that one participant scored at the 91st percentile on the BNT but at the 7th percentile on NAB naming Form 2.

Table 3.

Characteristics of participants obtaining lowest scores on NAB naming form 2

Participant BNT scaled score NAB naming form 1 T-score NAB naming form 2 T-score Age Years of education 
13 45 38 69 14 
35 27 79 14 
10 40 33 87 12 
14 41 35 79 15 
Participant BNT scaled score NAB naming form 1 T-score NAB naming form 2 T-score Age Years of education 
13 45 38 69 14 
35 27 79 14 
10 40 33 87 12 
14 41 35 79 15 

Note: BNT = Boston Naming Test; NAB = Neuropsychological Assessment Battery.

Discussion

This study investigated the validity of the naming test from the NAB by exploring its convergent validity and divergent validity. This represents an independent demonstration of the measure's validity and alternate forms reliability, in addition to demonstration by the creators of the test (White & Stern, 2003). Convergent validity was established by assessing the NAB naming test's correlation with an established measure of word-finding, the BNT. The NAB naming test was more correlated with this measure than with any other measures in a comprehensive neuropsychological test battery. Alternate forms reliability (r = .45) was also found between Forms 1 and 2 of the NAB naming test. This is different than the value found by White and Stern (2003; r = .72). It is likely that White and Stern utilized a Pearson correlation rather than Spearman's ρ; in our sample, the Pearson correlation coefficient was .74, more similar to the value found by White and Stern, but Spearman's ρ was considered a more appropriate correlation coefficient to use because of the non-normality of our sample.

A careful review of Table 1 reveals moderate differences between the normative scores on the NAB naming tests and the BNT. The mean scores on the NAB naming tests are both +0.5 SD above the mean, whereas the mean BNT score is +1.0 SD above the mean. In addition, the ranges of standard scores on both measures are notably different. Specifically, the range of standard scores on the NAB naming is from −1.5 SD to +1.5 SD around the mean, but the range on the BNT is from −0.7 SD to +2.2 SD around the mean. A single person could obtain a score of −1.5 SDs below the mean on the NAB naming but −0.7 SDs below the mean on the BNT and receive different interpretations of their performance (i.e., impaired vs. average). This suggests that clinicians should be cautious in their interpretations of performance on the two naming tests.

Convergent validity was also explored by assessing the NAB naming test's correlations with measures of memory, constructs that are related but also distinct from word-finding ability. White and Stern (2003) conducted correlations between the NAB naming test and measures of memory from the same test battery (i.e., the NAB), but did not explore correlations between the NAB naming test and already-established measures of memory (e.g., the California verbal learning test). The size of the correlation between the NAB naming test and the NAB memory index score (which incorporates both verbal and visual memory tests) was .38 in their sample, similar to the correlations ranging from .23 to .35 found with measures of memory in our sample. Thus, it can be said that both forms of the NAB naming test correlate moderately with measures of memory to a similar degree as the BNT correlated with them in our sample. This was expected, as all the naming tests involve memory for the names of objects.

It was expected that visuospatial skills would be related to the naming tests because of the nature of this test, in which patients look at a picture and name it. White and Stern (2003) found the naming test to correlate .38 with the spatial index score from the NAB (compared with correlations of .35 and .20 found with the judgment of line orientation in our sample), but they did not explore the degree to which the BNT correlates with visuospatial measures. Although it would be ideal for the naming tests to be independent of visuoperceptual skills, the nature of the test requires these skills. The test's validity as a measure of word-finding is enhanced, however, as the demand on visuoperceptual skills is lowered. Both forms of the naming test were correlated with a measure of visuoperceptual skills, and the BNT was more highly related to this measure (r = .49). Indeed, the strength of this correlation is similar to that of the BNT's correlation with the NAB naming tests (r = .45 and r = .50). These findings suggest that visuoperceptual skills may be less necessary to complete the NAB naming tests than to complete the BNT. The color photographs of the NAB naming test, as opposed to the black and white line drawings of the BNT, likely account for this.

White and Stern (2003) report the proportion of variance in NAB naming test scores attributable to demographic factors. We conducted similar analyses and also conducted these analyses with the BNT. Results suggest that although the BNT is highly related to educational level, the NAB naming tests are to a slightly lesser degree (though the difference in size of correlations was not statistically significant). These findings for the BNT are consistent with past research (Hawkins & Bender, 2002; Ross & Lichtenberg, 1998; Zec et al., 2007). Future research should investigate the relationships between the NAB naming test and education in other samples, particularly samples with fewer years of education. The BNT was also correlated with sex, with men outperforming women, whereas neither form of the NAB naming test was. This may have been due to sex being correlated with education in our sample, with men having more years of education. The NAB naming tests were correlated with age, indicating the need for age-based norms.

The NAB naming tests were not found to be correlated with measures of processing speed and abstraction, which were considered to be related to word-finding to only a small degree, if at all. This establishes divergent, or discriminant, validity; that is, it provides evidence that the NAB naming tests tap into a construct distinct from these other abilities. White and Stern (2003) did not discuss divergent validity in their technical manual.

This study was conducted with a predominantly European American sample who was well-educated (mean years of education = 15.1) and physically healthy. Future research should investigate the NAB naming test's validity in other ethnic groups, groups with less exposure to American culture, and groups with fewer years of education. This study should also be replicated in a clinical sample, as this study was conducted with healthy, community-dwelling volunteers. Lastly, a weakness of both the BNT and NAB naming test is that they require intact visual perception. As loss of vision is the third most common physical impairment among adults over 65 years of age (Morello & Fox, 1989), it would be ideal if there existed a well-established measure of naming ability that did not require visual perceptual skills.

Neuropsychologists attempt to utilize measures that are closer to being measures of the true construct of interest (i.e., word-finding ability) than of variables impacting performance on these measures (e.g., visuospatial perception, education, gender). Results suggest that the NAB naming test is a valid measure of word-finding ability that is not highly related to extraneous variables of less interest to the clinician.

Funding

Funding for this project was provided by the University of Colorado at Colorado Springs (B.P.Y).

Conflict of Interest

None declared.

Acknowledgements

The authors wish to thank Kelli Klebe and Peter Lichtenberg for their invaluable assistance with this manuscript.

References

Baldo
J. V.
Delis
D. C.
Wilkins
D. P.
Shimamura
A. P.
Is it bigger than a breadbox? Performance of patients with prefrontal lesions on a new executive function test
Archives of Clinical Neuropsychology
 , 
2004
, vol. 
19
 (pg. 
407
-
419
)
Benedict
R.
Brief visuospatial memory test-revised: Professional manual.
 , 
1997
Lutz, FL
Psychological Assessment Resources, Inc
Benedict
R.
Schretlen
D.
Groninger
L.
Dobraski
M.
Shpritz
B.
Revision of the brief visuospatial memory test: Studies of normal performance, reliability, and validity
Psychological Assessment
 , 
1996
, vol. 
8
 
2
(pg. 
145
-
153
)
Benton
A. L.
Sivan
A. B.
Hamsher
K. deS
Varney
N. R.
Spreen
O.
Contributions to neuropsychological assessment
 , 
1994
2nd ed.
Orlando, FL
Psychological Assessment Resources, Inc
Benton
A. L.
Varney
N. R.
Hamsher
K.
Visuospatial judgment: A clinical test
Archives of Neurology
 , 
1978
, vol. 
35
 (pg. 
364
-
367
)
Blumenfeld
H.
Neuroanatomy through clinical cases
 , 
2002
Sunderland, MA
Sinauer Associates
Charlson
M.
Pompei
P.
Ales
K.
MacKenzie
C.
A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation
Journal of Chronic Diseases
 , 
1987
, vol. 
40
 
5
(pg. 
373
-
383
)
Delis
D.
Kaplan
E.
Kramer
J.
Delis–Kaplan executive function system: Examiner's manual
 , 
2001
San Antonio, TX
The Psychological Corporation
Delis
D.
Kaplan
E.
Kramer
J.
Delis–Kaplan executive function system (D-KEFS): Technical manual.
 , 
2001
San Antonio, TX
The Psychological Corporation
Delis
D.
Kramer
J.
Kaplan
E.
Ober
B.
California verbal learning test (2nd ed.): Manual
 , 
2000
San Antonio, TX
The Psychological Corporation
Fastenau
P. S.
Denburg
N. L.
Mauer
B. A.
Parallel short forms for the Boston naming test: Psychometric properties and norms for older adults
Journal of Clinical and Experimental Neuropsychology
 , 
1998
, vol. 
20
 (pg. 
828
-
834
)
Graves
R. E.
Bezeau
S. C.
Fogarty
J.
Blair
R.
Boston naming test short forms: A comparison of previous forms with new item response theory based forms
Journal of Clinical and Experimental Neuropsychology
 , 
2004
, vol. 
26
 (pg. 
891
-
902
)
Hawkins
K. A.
Bender
S.
Norms and the relationship of Boston naming test performance to vocabulary and education: A review
Aphasiology
 , 
2002
, vol. 
16
 
12
(pg. 
1143
-
1153
)
Henry
J. D.
Crawford
J. R.
Phillips
L. H.
Verbal fluency performance in dementia of the Alzheimer's type: A meta-analysis
Neuropsychologia
 , 
2004
, vol. 
42
 (pg. 
1212
-
1222
)
Hodges
J. R.
Patterson
K.
Graham
N.
Dawson
K.
Naming and knowing in dementia of Alzheimer's type
Brain and Language
 , 
1996
, vol. 
54
 (pg. 
302
-
325
)
Ivnik
R. J.
Malec
J. F.
Smith
G. E.
Tangalos
E. G.
Peterson
R. C.
Neuropsychological test norms above age 55: COWAT, BNT, MAE token, WRAT-R reading, AMNART, Stroop, TMT, and JLO
The Clinical Neuropsycyhologist
 , 
1996
, vol. 
10
 (pg. 
262
-
278
)
Kaplan
E. F.
Goodglass
H.
Weintraub
S.
The Boston naming test: Experimental edition.
 , 
1978
Boston
Kaplan & Goodglass
Kaplan
E. F.
Goodglass
H.
Weintraub
S.
The Boston naming test
 , 
2001
2nd ed.
Philadelphia
Lippincott Williams & Wilkins
Lansing
A. E.
Ivnik
R. J.
Cullum
C. M.
Randolph
C.
An empirically derived short form of the Boston naming test
Archives of Clinical Neuropsychology
 , 
1999
, vol. 
14
 (pg. 
481
-
487
)
Lichtenberg
P. A.
Mental health practice in geriatric health care settings.
 , 
1998
Binghamton, NY
Haworth Press
Lichtenberg
P. A.
Vangel
S. J.
Kimbarow
M. L.
Ross
T. P.
The Boston naming test: Clinical utility
Clinical Gerontologist
 , 
1996
, vol. 
16
 
4
(pg. 
69
-
72
)
Lucas
J. A.
Ivnik
R. J.
Smith
G. E.
Ferman
T. J.
Willis
F. B.
Peterson
R. C.
, et al.  . 
Mayo's Older African American Normative Studies: Norms for Boston naming test, controlled oral word association, category fluency, animal naming, token test, WRAT-3 reading, trail making test, stroop test, and judgment of line orientation
The Clinical Neuropsychologist
 , 
2005
, vol. 
19
 (pg. 
243
-
269
)
Mack
W. J.
Freed
D. M.
Williams
B. W.
Henderson
V. W.
Boston naming test: Shortened version for use in Alzheimer's disease
Journal of Gerontology
 , 
1992
, vol. 
47
 (pg. 
164
-
168
)
Manly
J. J.
Miller
S. W.
Heaton
R. K.
Byrd
D.
Reilly
J.
Velasquez
R. J.
, et al.  . 
The effect of African American acculturation on neuropsychological test performance in normal and HIV-positive individuals
Journal of the International Neuropsychological Society
 , 
1998
, vol. 
4
 (pg. 
291
-
302
)
Morello
R. F.
Fox
D.
The aging eye
Geriatric Medicine Today
 , 
1989
, vol. 
8
 (pg. 
83
-
87
)
Nebes
R. D.
Semantic memory in Alzheimer's disease
Psychological Bulletin
 , 
1989
, vol. 
106
 
3
(pg. 
377
-
394
)
Rabin
L. A.
Barr
W. B.
Burton
L. A.
Assessment practices of clinical neuropsychologists in the United States and Canada: A survey of INS, NAN, and APA Division 40 members
Archives of Clinical Neuropsychology
 , 
2005
, vol. 
20
 (pg. 
33
-
65
)
Reynolds
C. R.
Kamphaus
R. W.
Reynolds intellectual screening test.
 , 
2003
Lutz, FL
Psychological Assessment Resources, Inc
Riccio
C.
Hynd
G.
Validity of the Benton's judgment of line orientation test
Journal of Psychoeducational Assessment
 , 
1992
, vol. 
10
 (pg. 
210
-
218
)
Ross
T. P.
Lichtenberg
P. A.
Expanded normative data for the Boston naming test for use with urban, elderly medical patients
The Clinical Neuropsychologist
 , 
1998
, vol. 
12
 
4
(pg. 
475
-
481
)
Salmon
D. P.
Bondi
M. W.
Neuropsychological assessment of dementia
Annual Review of Psychology
 , 
2009
, vol. 
60
 (pg. 
257
-
282
)
Saxton
J.
Ratcliff
G.
Munro
C. A.
Coffey
C. E.
Becker
J. E.
Fried
L.
, et al.  . 
Normative data on the Boston naming test and two equivalent 30-item short-forms
The Clinical Neuropsychologist
 , 
2000
, vol. 
14
 (pg. 
526
-
534
)
Stern
R. A.
White
T.
Neuropsychological assessment battery: Administration, scoring, and interpretation manual.
 , 
2003
Lutz, FL
Psychological Assessment Resources, Inc
Strauss
E.
Sherman
E. M. S.
Spreen
O.
A compendium of neuropsychological tests: Administration, norms, and commentary
 , 
2006
3rd ed.
New York
Oxford University Press
Testa
J. A.
Ivnik
R. J.
Boeve
B.
Pedersen
R. C.
Pankratz
V. S.
Knopman
D.
, et al.  . 
Confrontation naming does not add incremental diagnostic utility in MCI and Alzheimer's disease
Journal of the International Neuropsychological Society
 , 
2004
, vol. 
10
 (pg. 
504
-
512
)
The Psychological Corporation
Wechsler adult intelligence scale (3rd ed.)
 , 
1997
San Antonio
Author
Thompson-Schill
S. L.
Gabrieli
J. D.
Fleischman
D. A.
Effects of structural similarity and name frequency on picture naming in Alzheimer's disease
Journal of the International Neuropsychological Society
 , 
1999
, vol. 
5
 (pg. 
659
-
667
)
White
T.
Stern
R. A.
Neuropsychological assessment battery: Psychometric and technical manual.
 , 
2003
Lutz, FL
Psychological Assessment Resources, Inc
Williams
B. W.
Mack
W.
Henderson
V. W.
Boston naming test in Alzheimer's disease
Neuropsychologia
 , 
1989
, vol. 
27
 
8
(pg. 
1073
-
1079
)
Williams
V. G.
Bruce
J. M.
Westervelt
H. J.
Davis
J. D.
Grace
J.
Malloy
P. F.
, et al.  . 
Boston naming performance distinguishes between Lewy body and Alzheimer's dementias
Archives of Clinical Neuropsychology
 , 
2007
, vol. 
22
 (pg. 
925
-
931
)
Zec
R. F.
Burkett
N. R.
Markwell
S. J.
Larsen
D. L.
A cross-sectional study of the effects of age, education, and gender on the Boston Naming test
The Clinical Neuropsychologist
 , 
2007
, vol. 
21
 (pg. 
587
-
616
)