Content uploaded by Lucy Tamsin Dipper
Author content
All content in this area was uploaded by Lucy Tamsin Dipper on Mar 07, 2014
Content may be subject to copyright.
This article was downloaded by: [City University London]
On: 16 March 2012, At: 06:08
Publisher: Psychology Press
Informa Ltd Registered in England and Wales Registered Number: 1072954
Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Aphasiology
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/paph20
The Boston Naming Test in Greek:
Normative data and the effects of
age and education on naming
Aggeliki Patricacou
a
, Eirini Psallida
a
, Tim Pring
a
& Lucy
Dipper
a
a
City University, London, UK
Available online: 26 Nov 2007
To cite this article: Aggeliki Patricacou, Eirini Psallida, Tim Pring & Lucy Dipper (2007): The
Boston Naming Test in Greek: Normative data and the effects of age and education on naming,
Aphasiology, 21:12, 1157-1170
To link to this article: http://dx.doi.org/10.1080/02687030600670643
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-
conditions
This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any
representation that the contents will be complete or accurate or up to date. The
accuracy of any instructions, formulae, and drug doses should be independently
verified with primary sources. The publisher shall not be liable for any loss, actions,
claims, proceedings, demand, or costs or damages whatsoever or howsoever caused
arising directly or indirectly in connection with or arising out of the use of this
material.
#
2007 Psychology Press, an imprint of the Taylor & Francis Group, an Informa business
http://www.psypress.com/aphasiology DOI: 10.1080/02687030600670643
The Boston Naming Test in Greek: Normative data and the
effects of age and education on naming
Aggeliki Patricacou, Eirini Psallida, Tim Pring and Lucy Dipper
City University, London, UK
Background: The Boston Naming Test (BNT) is widely used as a clinical assessment of
language and cognitive deficits. It has been adapted and translated for use in other
languages and cultures.
Aims: This study translated and adapted the test for use in Greece. Normative data were
collected on the test for healthy Greek speakers of different ages and educational
backgrounds.
Methods and Procedures: Participants in four different age ranges and with three levels
of educational achievement were tested. They were screened for cognitive decline using a
Greek version of the mini mental state examination.
Outcomes and Results: Strong effects of age and education were found on naming. The
former replicates previous results. Results on the latter have been less consistent and
their occurrence here reflects the greater inequality in educational opportunity that has
existed in Greece until comparatively recent times. Significant interactions between age,
education, and gender are interpreted as reflecting changing social and gender roles in
Greek society. A reordering of items reflecting their difficulty for this Greek sample is
presented for clinical use.
Conclusions: This study provides norms for a Greek version of the BNT. These highlight
the effects of age and education on naming. Scores for many older and less-educated
participants might be taken to indicate pathology despite their lack of neurological or
cognitive problems. This illustrates the need for norms that reflect local circumstances
and the need to update norms as social and educational changes occur.
The Boston Naming Test (BNT) is a confrontation naming test that is widely used
clinically to detect mild word-retrieval problems in aphasia, brain injury, and
dementia. The test exists in several forms. Borod, Goodglass, and Kaplan (1980)
collected normative data on an initial version of the test consisting of 85 drawings.
Kaplan, Goodglass, and Weintraub (1983) introduced the more familiar 60-item test,
which is now the most widely used version. Subsequently shorter versions with 30 or
15 items have been developed to facilitate the testing of clients with neurological
impairments (Lansing, Ivnik, Cullum, & Randolph, 1999; Mack, Freed, Williams, &
Henderson, 1992; Williams, Mack, & Henderson, 1989).
The BNT is most widely used to assess word-finding problems in aphasia but has
been used with many other client populations (see Marien, Mampaey, Vervaet,
Saerens, & DeDeyn, 1998, for a summary). A number of studies have provided
Address correspondence to: Dr Tim Pring, Department of Language and Communication Science,
City University, Northampton Square, London, UK. E-mail: T.R.Pring@city.ac.uk
APHASIOLOGY, 2007, 21 (12), 1157–1170
Downloaded by [City University London] at 06:08 16 March 2012
normative data on normally functioning individuals. As well as providing
comparison data for impaired performance, these studies have investigated the
effects of demographic variables such as age, education, and gender on naming
performance. The popularity of the BNT has also led to the collection of normative
data from English-speaking participants in other countries and to the translation of
the test into several other languages. Translations of the test include versions in
French (Colombo & Assal, 1992), Spanish (Allegri, Mangone, Villavicencio,
Rymberg, & Baumann, 1997; Quinones-Ubeda, Pena-Casanova, Bohm, Gramunt-
Fombuena, & Comas, 2004), Dutch (Marien et al., 1998), Korean (Kim & Na, 1999),
Chinese (Cheung, Cheung, & Chan, 2004), and Swedish (Tallberg, 2005). English
versions of the test have also been used in studies in Australia (Cruice, Worrall, &
Hickson, 2000; Worrall, Yiu, Hickson, & Barnett, 1995), New Zealand (Barker-
Collo, 2001), and Canada (Roberts, Garcia, & Desrochers, 2002)
The effects of age and education on naming have been widely investigated. The
results have been less consistent than might be expected. Two factors complicate the
outcome. First, the majority of studies are cross-sectional. As a result comparisons
across age groups may be subject to cohort effects. Second, age and education are
likely to be confounded. Older groups are likely to have fewer years of education and
may therefore show declines in naming that result from either age or education or
both. Early studies (Borod et al., 1980; Nicholas, Obler, Albert, & Goodglass, 1985)
found that naming remained consistent until about 70 years of age, after which it
declined. In contrast, Van Gorp, Satz, Klersch, and Henry (1986) found only a mild
decrease with age in groups of participants ranging from 60 to 80+. This result
should be interpreted cautiously, however. Fully representative samples are rarely
obtained (see Hawkins & Bender, 2002). Typically participants are drawn from the
healthy, active, and well educated. This problem increases with age. The proportion
of healthy active participants grows smaller but may be more accessible and more
easily tested. In Van Gorp et al., the mean verbal IQ of each age group is at least one
standard deviation above the population mean. The data for the elderly groups were
intentionally based on a small number of bright, independently living participants.
Schmitter-Edgecombe, Vesneski, and Jones (2000) obtained rather similar results. In
their study, groups of participants were matched for IQ, which was again above
average. The older groups obtained higher naming scores than younger participants,
but had had more years of education. These results run counter to the expectation
that scores will decline in older participants due both to age and to their relative lack
of educational experience.
The implication of these findings appears to be that naming need not decline with
age. However, confirmation that the samples in the studies by Van Gorp et al. and
Schmitter-Edgecombe et al. were unusual comes from studies of more diverse elderly
participants. Ross and colleagues (Lichtenberg, Ross, & Christensen, 1994; Ross &
Lichtenberg, 1998; Ross, Lichtenberg, & Christensen, 1995) obtained much lower
scores for elderly clients who had more varied medical and educational backgrounds.
Others have also found a decline in naming with age (La Barge, Edwards, &
Knesevich, 1986; Marien et al., 1998; Welch, Doineau, Johnson, & King, 1996;
Worrall et al., 1995). However, two related reservations about this seemingly
straightforward result should be noted. The first concerns the magnifying effect that
cohort effects may have in cross-sectional studies. Only a few studies have examined
naming over time and only for comparatively short intervals. Conflicting results are
available. Cruice et al. (2000) found that age and naming were significantly
1158
PATRICACOU ET AL.
Downloaded by [City University London] at 06:08 16 March 2012
correlated in their cross-sectional data but found no decline over 4 years in their
longitudinal data. Connor, Spiro, Obler, and Albert (2004) also compared the two
approaches and found that cross-sectional data increased the apparent effect of age.
In contrast, Au, Joung, Nicholas, Obler, Kass, and Albert (1995), who retested their
participants over a 7-year period, found a decline that was more pronounced in older
participants. A second reservation concerns the role of confounding variables in
simulating or enhancing age effects. Tallberg (2005), who examined a range of
variables, found that education and measures of cognitive ability and verbal fluency
were significantly correlated with naming, but age was not. This result echoes those
above where little effect of age was found in healthy, well-educated participants, and
suggests that other variables must be acknowledged in assessing naming ability in
elderly participants.
Similar inconsistent findings exist for the effect of education on naming perfor-
mance. Several studies have found a relationship (Allegri et al., 1997; Henderson,
Frank, Pigatt, Abramson, & Houston, 1998; Kim & Na, 1999; Marien et al., 1998;
Tallberg, 2005; Worrall et al., 1995). Welch et al. (1996) found an interaction between
age and education indicating that elderly participants with low educational levels
scored particularly poorly. Rosselli and Ardila (1990) argue that participants with low
educational level can function adequately in everyday life but perform poorly when
examined on neuropsychological tests, and Neils, Baris, Carter, Dell’aria, Nordloh,
Weiler, et al. (1995) stress the need for clinicians to know the role played by education
when using the BNT. However, other studies have failed to find an effect of education
(Farmer, 1990; Fastenau, Denburg, & Mauer, 1998; Nicholas, Brookshire, MacLennan,
Schumacher, & Porrazzo, 1989; Randolph, Lansing, Ivnik, Cullum, & Hermann,
1999; Tombaugh & Hubley,1997; Van Gorp et al., 1986). Hawkins and Bender (2002)
have contested these null findings, arguing that the studies involved had an insufficient
range of educational experience. Several of the positive findings above may owe
something to sampling populations in countries where a greater range of educational
opportunity exists. Caution is required with such samples, however, as educational
opportunity is likely to have increased with time, leading to strong cohort effects.
Hawkins and colleagues (Hawkins & Bender, 2002; Hawkins, Sledge, Orleans,
Quinlan, Rakfeldt, & Hoffman, 1993,) also argue that vocabulary size is a better
predictor of BNT performance than education. They and others (Thompson &
Heaton, 1989) have found correlations in the vicinity of 0.80 between measures of
vocabulary size and the BNT, whereas education typically gives correlations between
0.50 and 0.60. However, this relationship may not be useful for clinical purposes,
where premorbid vocabulary size will be difficult to estimate. Here current test
performance should be compared against data for unimpaired participants with a
similar level of education.
There has been less interest in the effect of gender on BNT performance. Most
studies have failed to find an effect (Henderson et al., 1998; LaBarge et al., 1986;
Ross et al., 1995). Others have found an interaction between age and gender (Marien
et al., 1998; Randolph et al., 1999; Welch et al., 1996). In these a male advantage for
naming is found in older participants and is likely to reflect the greater educational
opportunities experienced by males than females in these older populations.
The BNT was developed for a North American population. Consequently,
adaptations of it for use in other countries and languages are likely to reflect cultural
differences. Worrall et al. (1995) found that Australians obtained lower scores on the
BNT, which they attributed to a cultural bias in the test. Biases arise because items
THE BOSTON NAMING TEST IN GREEK 1159
Downloaded by [City University London] at 06:08 16 March 2012
on the test differ in their familiarity in different cultures. In some cases, items may be
virtually unknown in the new culture. Worrall et al. (1995) found the items ‘‘pretzel’’
and ‘‘beaver’’ problematic and replaced them with ‘‘pizza’’ and ‘‘platypus’’.
Although this had little effect on the scores, the new items were thought to be
more acceptable to Australians. A cultural bias was also found by Barker-Collo
(2001) who tested university students in New Zealand. Scores were lower than North
American samples of a similar age. She identified six items on which New Zealanders
had high error rates, and suggested that the test be adapted by replacing these. Five
of the items also had high error rates in Worrall et al.’s Australian data.
Where the test is translated, linguistic differences add to those due to culture. Item
names may not translate directly into the new language or may be more likely to
result in the production of synonyms, thus complicating the scoring. There will be
the dilemma seen above as to whether to continue with the original test or to replace
some items that are inappropriate for cultural and/or linguistic reasons. Kim and Na
(1999) adopted an extreme approach to this problem in their development of a
Korean version of the test. They pilot tested a large number of items before selecting
a set of 60 to form a parallel version of the test. The resulting test included only 10 of
the original items.
Replacing items serves to make scores a fairer reflection of naming ability and to
improve the face validity of the test in the new country/language. It may not be
straightforward, however. Several guidelines might be followed. It is desirable that
the replacement item is from the same semantic category as the item replaced so as to
maintain the semantic diversity of the test. Replacement items should also match the
difficulty of the item in the original test. This might be achieved by choosing items
with a similar frequency to the replaced items in either English or the new language.
The latter may prove difficult if frequency information is not available in the new
language (although this is not the case for Greek) or if the frequency counts in the
different languages have used different methodologies or searched different materials
to obtain their information.
In the original BNT, items are placed in the order of respondents’ naming
accuracy rather than frequency. Authors of versions of the test in different languages
have also used naming accuracy as a measure of difficulty and have either presented
these data or suggested a reordering of the test items based on them (see Allegri et al.,
1997; Marien et al., 1998; Roberts et al., 2002; Tallberg, 2005). Evidence in favour of
this approach may be obtained by examining the frequencies of items on the BNT.
The mean frequency is 21.37 per million words (Francis & Kucera, 1982). In contrast
the median frequency is 2, reflecting a marked positive skew in the frequencies of the
items. Only 10 items, mainly occurring early in the test, have frequencies greater than
10 per million. The remaining items are low in frequency, and their position in the
test, determined by the difficulty respondents had in naming them, bears little
relationship to the small differences in their frequency. It appears that factors other
than frequency determine the ease of naming of many of these items. It also suggests
that the suitability of replacement items in new versions of the test may be judged by
examining the difficulty respondents have in naming them (and therefore their
position in the test), which should equate with the difficulty of the replaced items.
This policy was used in the present paper, which reports normative data on a Greek
language version of the BNT. A pilot test was used to identify items that were
unfamiliar and four were replaced by low-frequency items in Greek whose naming
difficulty was compared with that of the omitted items.
1160
PATRICACOU ET AL.
Downloaded by [City University London] at 06:08 16 March 2012
Norms were collected for people of different ages, genders, and years of
education. Care was taken to find people at each level of education and for each age
sampled so that the effect of both variables could be assessed and cohort effects
might be avoided. The role of gender and education are also of interest given
historical changes in Greek society. Educational opportunity has substantially
changed within the lifetime of the older participants tested and differences also exist
between urban and rural sections of the population. These differences may also
interact with gender. Different educational opportunity and different roles within
society have traditionally existed and may be particularly apparent in the older
group of participants tested.
METHOD AND PROCEDURE
Participants
A total of 100 individuals were tested. All were native Greek speakers selected to
represent different ages and levels of education. Half were male and half female.
They represented four age groups: 20–40, 41–60, 61–70, and 71+ and three levels of
education, 0–6, 7–12, and 13+ years of attendance in full-time education. Smaller age
bands were used for older participants where naming might be expected to decline.
All participants were living independently in the community. They were recruited
through social organisations, groups for the elderly, universities, factories,
companies, and some by word of mouth. All were volunteers. None had a reported
history of neurological or cognitive disorder.
Pilot study
A pilot study using the original BNT items was carried out to identify items that are
culturally specific. Ten participants aged from 30 to 40 participated. Four were
female and six male, and they had 13+ years of education. They were selected
because their age and education suggested that they would perform well on the test.
Consequently, items that they found difficult would be likely to trouble other
participants. Eight or more of these participants gave wrong responses on four of the
items. These were:
(a) Pretzel: Eight participants gave the category name (cookie) instead of the
specific name.
(b) Doorknocker: Only two participants gave the correct name roptro.
(c) Stethoscope: Nine gave the name ‘‘medical hearings’’ (iatrika akoustica) rather
than stethoscopio.
(d) Scroll: All participants gave the word ‘‘papyrus’’. The exact translation is rolo.
New items from the same categories were chosen to replace these items:
(a) ‘‘tsureki’’, which is a kind of cake, replaced pretzel;
(b) ‘‘grammatokivotio’’ (mailbox) replaced ‘‘doorknocker’’;
(c) ‘‘piesometro’’, an instrument that measures blood pressure, replaced ‘‘stetho-
scope’’;
(d) ‘‘kionas’’, an ancient Greek column, replaced ‘‘scroll’’.
THE BOSTON NAMING TEST IN GREEK 1161
Downloaded by [City University London] at 06:08 16 March 2012
The reasons for the poor performance on these items appear to be varied. Pretzels
are unknown in Greece and ‘‘roptro’’ is from ancient Greek and rarely used.
However, ‘‘stethoscopio’’ and ‘‘rolo’’ are not particularly rare and errors here may
have more to do with the appropriateness of the pictures. The replacement items are
all of low frequency in Greek. A word count of 34 million words (Institute for
Language and Speech Processing, 2000) shows that each occurs less than
once per million words.
Black and white line drawings of similar style and size to the BNT drawings were
used for the replacement items. The items and their drawings were tested for cultural
appropriateness. Ten further participants were shown the pictures and all were able
to recognise them.
The pilot study also found that synonyms were produced for several items. They
were particularly common for Wheelchair, Latch, Tongs, and Globe. Such responses
are unacceptable if standard scoring guidelines are followed. Other authors have
argued that frequently elicited synonyms should be counted as correct (Cruice et al.,
2000; Nicholas et al., 1989; Worrall et al., 1995). This policy was adopted for the four
items above.
Administration
Before testing participants answered a questionnaire by giving information on their
age, sex, medical history (neurological and cognitive problems), education,
occupation, place of residence, and interests and hobbies. Data from this were used
to assign the participants to the groups described above.
All participants did the Greek Mini-Mental State Examination (Fountoulakis,
Tsolaki, Chantzi, & Kazis, 1994). Participants who reported a history of
neurological illness or who had scores lower than 23 were excluded.
Participants were tested individually in a quiet room by one of two qualified
speech and language therapists. Both followed the same administration and scoring
rules. Standard BNT administration for non-aphasic people starts with item 30, as
no failures are expected before this, and is discontinued following six consecutive
failures. Here all 60 pictures were tested in order to determine their order of difficulty
in Greek, which might differ from that of the original test.
Correctly named items were recorded and the examiner proceeded to the next
item. Spontaneous self-corrections were given credit. If no answer was given during
20 seconds the examiner gave the semantic cue and correct answers within 20 seconds
were given credit. After responses indicating a misperception—e.g., ‘‘umbrella’’ for
‘‘mushroom’’—a semantic cue was provided and credit given if the target was
produced. If a semantically related word was produced, an eliciting cue was given
(Can you think of another word?) and credit given if the exact word was produced
(Cruice et al., 2000). Lenient scoring was used for the four items for which synonyms
were given in the pilot study. Strict scoring applied for the rest of the test. The overall
score was calculated by adding correct spontaneous responses and those following a
stimulus cue.
Incorrect responses were transcribed orthographically. Administration of the test
was recorded on audiotape to evaluate the reliability of the coding and scoring.
Incorrect responses were coded using a procedure adapted from other authors
(Tombaugh & Hubley, 1997; Worrall et al., 1995). Responses were put into eight
categories: don’t know, misperceptions, semantic errors, perseveration, description,
1162
PATRICACOU ET AL.
Downloaded by [City University London] at 06:08 16 March 2012
synonyms, unrelated responses, and phonological errors. The two examiners carried
out the coding. Ambiguous responses were coded after they reached agreement. The
error types, their definitions, and some examples are given in Table 1.
RESULTS
The data were analysed with a three-factor between-participants analysis of variance
(ANOVA). Age had four levels (20–40, 41–60, 61–70, and 71+), education had three
levels (0–6, 7–12, and 13+ years), and gender two levels. The mean for the 100
participants was 42.86 (SD 9.76, range 15–58). This is lower than in other studies. In
Nicholas et al. (1989) the overall mean was 54.5; in Worrall et al. (1995) it was 52.76;
and in Van Gorp et al. (1986) it was 54.01.
The analysis showed that the main effects of age, F(3, 76) 5 18.03, p , .0001, and of
education, F(2, 76) 5 45.67, p , .0001, were significant. There was no main effect of
gender.
The means for each level of age are given in Table 2. Naming clearly declines with
age. A Newman-Keuls unplanned comparison was used to test for differences
TABLE 1
Types and examples of errors
Error types Definition and examples
Don’t know No attempt to name or describe object (‘‘don’t know’’, ‘‘don’t remember’’, ‘‘I
knew it but have forgotten’’)
Misperception Response indicating a visual misperception of the item. (‘‘frying-pan’’ for
racquet, ‘‘table’’ for mushroom)
Semantic Response that is the category name (‘‘plant’’ for cactus) or a related member
of the category (‘‘hippopotamus’’ for rhinoceros). Response associated with
the context or particular physical features of the item (‘‘cave’’ for igloo,
‘‘branch’’ for asparagus, ‘‘fire’’ for volcano, ‘‘rope’’ for noose)
Perseveration Incorrect repetition of a previous response (‘‘compass’’ for protractor)
Description Response giving information about the characteristics or use of the object
(‘‘painters use it’’ for palette, ‘‘it’s made out of wood for the flowers to climb
on’’ for trellis)
Synonym Response with the same meaning as the target (‘‘kolona’’ for kionas, both
meaning column)
Unrelated Response with no relationship to the target (‘‘bed’’ for harp)
Phonological Mispronounced words sharing at least one common syllable (‘‘liomenes
scales’’ for kiliomenes scales)
TABLE 2
Means for different age groups
Age N Means SD Range
21–40 27 48.89 6.22 33–57
41–60 31 45.87 7.84 32–58
61–70 21 39.43 8.41 18–49
71+ 21 34.13 10.14 15–49
THE BOSTON NAMING TEST IN GREEK
1163
Downloaded by [City University London] at 06:08 16 March 2012
between the age groups. Naming declined significantly at each age. The 71+ group
were significantly poorer than each of the other groups (p , .01). The 61–70 group
were significantly poorer than the 41–60 and 21–40 groups (p , .01) and the 41–60
group were significantly poorer than the 21–40 age group (p , .05)
The means for each level of education are given in Table 3. Naming increases with
years in education. A Newman-Keuls unplanned comparison found that each level
of education differed from each other level (p , .01).
Significant interactions were found between education and gender, F(2, 76) 5
7.02, p , .01, and between age, education and gender, F(6, 76) 5 2.41, p , .05.
Means for the former are given in Table 4. The p values are from an analysis of
simple main effects. The results show that both males and females increased their
scores with education. Females show the more dramatic increase. At the 0–6 years
education level they were significantly poorer than males (p , .01) but at the 13+
level they are significantly better (p , .05).
Means for each group in the three-way interaction between age, education, and
gender are shown in Table 5. Again the p values were obtained via an analysis of
simple main effects.
This three-way interaction qualifies the two-way interaction above. It shows that
the trend for females to overtake males as they acquire education is strongest in the
older participants. In the youngest age group there are no differences between males
and females. A non-significant trend is seen in the 41–60 age group and this becomes
stronger in the older groups where gender differences with education become marked
and significant.
Error analysis
The overall accuracy of participants on the test was 71.43% (4287 items were correct
and 1713 incorrect). Appendix A gives the items in their order of difficulty (%
correct) in Greek. A number of individual items were found to have higher or lower
% accuracies than would be expected from their position in the test. Items that
proved difficult (with accuracies below 40%) were, ‘‘canoe’’ (35 %), ‘‘beaver’’ (28 %),
TABLE 3
Means for the three levels of education
Education level N Means SD Range
0–6 years 26 33.92 8.48 15–50
7–12 years 36 43.25 8.97 17–57
13+ years 38 48.63 6.29 30–58
TABLE 4
Means and p values for the education by gender interaction
0–6 7–12 13+ p value
Males 37.73 41.65 46.47 p,.001
Females 31.13 45.25 50.79 p,.001
p value p,.01 p5.06 p,.05
1164 PATRICACOU ET AL.
Downloaded by [City University London] at 06:08 16 March 2012
‘‘igloo’’ (39 %), ‘‘dominoes’’ (35%), ‘‘unicorn’’ (18%), ‘‘yoke’’ (27 %), ‘‘trellis’’ (26
%), ‘‘palette’’ (37 %), ‘‘abacus’’ (38%).
Appendix A also allows us to judge the appropriateness of the replacement items.
The four replaced items occurred in positions 19, 40, 42, and 53 (mean 5 38.5) of the
original test. The replacement items occur in positions 18, 27, 43, and 51
(mean534.75) of the Greek version, suggesting that their relative difficulty is
similar to that of the replaced items.
Incorrect responses were coded in the eight categories described above. When
multiple errors occurred to the same item, each was coded separately. Table 6 shows
the number and percentage of errors in each category.
The table shows that just over two-thirds of errors were semantic or description
errors. These errors reflect correct recognition of target items whose names were
either not known or not retrieved. In contrast only 11.64% of errors were to items
unknown to the participant. Other categories accounted for a small percentage of the
errors and some could be accounted for by idiosyncrasies within the test. For
example, perseveration errors occurred only when ‘‘tripod’’ (item 52) was incorrectly
TABLE 5
Means and p values for the age by education by gender interaction
0–6 7–12 13+ p value
Age 21–40 Males 37.50 52.00 52.40 p,.01
Females 38.33 49.20 51.71 p,.01
p value ns ns ns
Age 41–60 Males 40.67 47.40 50.17 p5.06
Females 34.80 43.83 54.14 p,.0001
p value ns ns ns
Age 61–70 Males 38.25 39.33 43.33 ns
Females 29.00 42.67 45.75 p,001
p value p,.05 ns ns
Age 71+ Males 32.50 31.14 38.00 ns
Females 20.67 43.50 47.50 p,.0001
p value p,.05 p,.05 p,.05
TABLE 6
Number and percentage of error types
Error type Number Percentage
Don’t know 246 11.64
Misperception 246 11.64
Semantic 1025 48.50
Description 391 18.50
Synonym 105 4.97
Perseveration 14 0.66
Unrelated 65 3.08
Phonological 24 1.14
THE BOSTON NAMING TEST IN GREEK
1165
Downloaded by [City University London] at 06:08 16 March 2012
named ‘‘compass’’ (item 50) and the majority of the phonological errors were on
‘‘abacus’’ (/arihmitirio/wrongly answered as /arihmometirio/). Many unrelated
responses occurred on ‘‘latch’’ (9), and on ‘‘harp’’, ‘‘yoke’’, and ‘‘column’’ (5 each),
and ‘‘column’’, ‘‘mailbox’’, and ‘‘escalators’’ accounted for most of the synonym
errors. Many misperception errors appeared to just three items: ‘‘igloo’’ (55), ‘‘dart’’
(28), and ‘‘volcano’’ (20), and may reflect ambiguously drawn pictures and
differences in cultural background rather than visual errors.
DISCUSSION
The Boston Naming Test is a significant tool that is widely used for the detection of
naming deficits in aphasia (Kohn & Goodglass, 1985) and in primary neurodegen-
erative disorders such as dementia (Williams et al., 1989). The test was designed for
use with English-speaking North American clients and normative data are available
on this population. Recognition of its value is reflected in the increasing number of
translations and adaptations of the test for use in several other languages and
cultures. The present paper adapts the BNT for use with Greek speakers, and
presents norms for participants at different ages and with different levels of
education.
Significant main effects of age and education were found, and significant
interactions between education, age, and gender. The decline in naming with age is
consistent with many previous studies. The strong effect of education on naming is
consistent with some earlier results (Neils et al., 1995; Nicholas et al., 1989; Ross
et al., 1995; Welch et al., 1996; Worrall et al., 1995) but not others (Farmer, 1990;
Fastenau et al., 1998; Nicholas et al., 1989; Randolph et al., 1999; Tombaugh &
Hubley, 1997; Van Gorp et al., 1986). However, the latter sampled populations
where high levels of educational attainment are the norm (the mean years of
education was 14.45 in Van Gorp et al. and 14.62 years in Farmer). In this respect
the Greek population provides a strong test of the effects of education, as large
inequalities existed in comparatively recent times and will have influenced the
performance of participants in this study.
Previous studies have sometimes (Marien et al., 1998; Welch et al., 1996) but not
always (La Barge et al., 1986; Ross et al., 1995) found effects of gender on naming.
No main effect of gender was found in this study. However, there were significant
interactions between education and gender, and education, age, and gender. The
former shows that males perform better than females at low levels of education but
more poorly at higher levels of education. Females, it appears, benefit more from
education. It is likely that other factors are influencing performance here. The older
participants in this study attended school at a time when many more males than
females were educated beyond the elementary level. Girls who did continue may have
been particularly able and better placed to benefit from education. This conclusion is
supported by the three-way interaction between education, age, and gender. No
gender differences were found in the younger age groups. These participants
attended school when gender biases in education were less pronounced. Here both
sexes increased their scores with education but did not differ. The interaction
between education and gender becomes progressively stronger with age. This is
consistent with the suggestion that few, and perhaps only the most able, females
progressed to higher levels of education at this time. Their scarcity was reflected in
the difficulty in finding them for the present study. Participants in the older age
1166
PATRICACOU ET AL.
Downloaded by [City University London] at 06:08 16 March 2012
groups with high levels of education were generally hard to locate; female
participants particularly so.
This explanation cannot account for the better naming of older male than female
participants at the lowest level of education. This may reflect the role of women in
Greek society during the lifetime of these older participants. Just as access to
education was limited, so their social role was largely restricted to caring for the
family, while males worked and may have experienced a more varied social life. The
changes in Greek society have been similar to, but more dramatic than, those in
other countries and have had a stronger effect on the norms in this Greek adaptation
of the BNT. They also suggest that norms will change over time and will need to be
reassessed in countries where educational opportunities and social roles change.
Greek participants in this study obtained lower scores than participants in other
studies. Here the overall mean score was 42.86. In the studies by Nicholas et al.
(1989), Van Gorp (1986), Farmer (1990), and Worrall et al. (1995) means of over 50
were obtained. This lower score is likely to reflect linguistic and cultural differences.
One reason for the lower scores concerns the inclusion of a high proportion of
participants who are expected to score poorly. Other studies have found it difficult if
not impossible to locate participants with fewer than 6 years of education. Here a
quarter of the participants had fewer than 6 years of education. Here also 42% of the
participants were above 60 and a fifth were over 70. Scores for the younger more
highly educated groups were in excess of 50 and closer to those found in other
studies.
The possibility of cohort effects in previous papers has meant that they have been
unable to determine the relative effects of age and education. By separating out their
effects, the present study shows that both have strong effects. It suggests that
normative data on the test (in all languages) should include data on different levels of
education as well as age, if the test is to be more effectively used clinically with
elderly participants who may have more varied educational experience than current
generations.
Cultural differences in the familiarity of the items in the test are likely to be a
second reason for the lower scores than in other studies. Prior to testing, a pilot with
10 participants showed that four items were particularly difficult and these were
replaced. Some other items were clearly less familiar to Greek participants than to
those for whom the BNT was designed, and further items might reasonably have
been replaced. A decision was taken to retain items that were merely less familiar so
that the general character of the test would remain as similar to the original as
possible.
The original BNT presents items in ascending order of difficulty. The Greek order
of difficulty appears to differ. In several cases items that occurred early in the test led
to an unexpectedly high number of errors, and later items were easier than
anticipated. This suggests that replacing items in new versions of the test may be less
than straightforward. Appendix A presents the test in the order of difficulty
experienced by our Greek respondents. If the Greek version of the test is to be used
clinically, it may be appropriate to present the items in this order. The reordering
also allowed us to demonstrate that the difficulty of the new items was similar to
those they replaced
This study provides a Greek version of the BNT, which may be useful to clinicians
working with Greek-speaking clients with linguistic or cognitive impairments. It also
highlights the effects of age and education on naming scores. Some of the older and
THE BOSTON NAMING TEST IN GREEK 1167
Downloaded by [City University London] at 06:08 16 March 2012
less-educated participants tested obtained scores that might normally be considered
to indicate pathology. Yet these participants had no history of neurological or
cognitive problems and had passed the cut-off score for entry on the mini-mental
state examination. If the explanation offered above is correct, the scores also
illustrate the extent to which changes in education and social circumstances influence
naming performance.
Manuscript received 8 September 2005
Manuscript accepted 2 March 2006
First published online 4 September 2007
REFERENCES
Allegri, R. F., Mangone, C. A., Villavicencio, A. F., Rymberg, S., & Baumann, D. (1997). Spanish Boston
Naming Test norms. The Clinical Neuropsychologist, 11, 416–420.
Au, R., Joung, P., Nicholas, M., Obler, L. K., Kass, R., & Albert, M. L. (1995). Naming ability across the
adult life span. Aging and Cognition, 2, 300–311.
Barker-Collo, S. L. (2001). Short report on the 60-Item Boston Naming Test: Cultural bias and possible
adaptations for New Zealand. Aphasiology, 15, 85–92.
Borod, J. C., Goodglass, H., & Kaplan, E. (1980). Normative data on the Boston Diagnostic Aphasia
Examination, Parietal Lobe Battery, and the Boston Naming Test. Journal of Clinical
Neuropsychology, 2, 209–215.
Cheung, R. W., Cheung, M. C., & Chan, A. S. (2004). Confrontation naming in Chinese patients with left,
right or bilateral brain damage. Journal of the International Neuropsychological Society, 10, 46–53.
Colombo, F. T., & Assal, G. (1992). Boston Naming Test – French Language adaptation and short forms.
European Review of Applied Psychology, 42, 67–73.
Connor, L. T., Spiro, A., Obler, A. K., & Albert, M. L. (2004). Change in object naming ability during
adulthood. Journals of Gerontology Series B–Psychological Sciences and Social Sciences, 59, 203–209.
Cruice, M. N., Worrall, L. E., & Hickson, L. M. H. (2000). Boston Naming Test results for healthy older
Australians. A longitudinal and cross sectional study. Aphasiology, 12, 143–155.
Farmer, A. (1990). Performance of normal males on the Boston Naming Test and The Word Test.
Aphasiology, 4, 293–296.
Fastenau, P. S., Denburg, N. L., & Mauer, B. A. (1998). Parallel short form for the Boston Naming Test:
Psychometric properties and norms for older adults. Journal of Clinical and Experimental
Neuropsychology, 20, 828–834.
Fountoulakis, K., Tsolaki, M., Chantzi, E., & Kazis, A. (1994). Mini Mental State Examination.
Validation study in demented patients from the elderly Greek population. Enkephalos (Greece), 31,
93–102.
Francis, W., & Kucera, H. (1982). Frequency analysis of English usage. Boston: Houghton Mifflin.
Hawkins, K. A., & Bender, S. (2002). Norms and the relationship of Boston Naming Test performance to
vocabulary and education: A review. Aphasiology, 16, 1143–1153.
Hawkins, K. A., Sledge, W. H., Orleans, J. F., Quinlan, D. M., Rakfeldt, J., & Hoffman, R. E. (1993).
Normative implications of the relationship between reading vocabulary and Boston Naming Test
performance. Archives of Clinical Neuropsychology, 8, 525–537.
Henderson, L. W., Frank, E. M., Pigatt, T., Abramson, R. K., & Houston, M. (1998). Race, gender, and
educational level effects on Boston Naming Test scores. Aphasiology, 12, 901–911.
Institute for Language and Speech Processing (2000). Athens. Retrieved from http://hnc.ilsp.gr 10/01/06.
Kaplan, E. F., Goodglass, H., & Wintraub, S. (1983). The Boston Naming Test. Experimental Edition.
Philadelphia: Lea & Febiger.
Kim, H. H., & Na, D. L. (1999). Normative data on the Korean version of the Boston Naming Test.
Journal of Clinical and Experimental Neuropsychology, 21, 127–133.
Kohn, S. E., & Goodglass, H. (1985). Picture-naming in aphasia. Brain and Language
, 24, 266–283.
La Barge, E., Edwards, D., & Knesevich, J. W. (1986). Notes and discussion, performance of normal
elderly on the Boston Naming Test. Brain and Language, 27, 380–384.
1168 PATRICACOU ET AL.
Downloaded by [City University London] at 06:08 16 March 2012
Lansing, A., Ivnik, R., Cullum, C. M., & Randolph, C. (1999). An empirically derived short form of the
Boston Naming Test. Archives of Clinical Neuropsychology, 14, 481–487.
Lichtenberg, P. A., Ross, T., & Christensen, B. (1994). Preliminary normative data on the Boston Naming
Test for an older urban population. Clinical Neuropsychologist, 8, 109–111.
Mack, W. J., Freed, D. M., Williams, B. W., & Henderson, V. W. (1992). Boston Naming Test: Shortened
version for use in Alzheimer’s disease. Journal of Gerontology, 47, 154–158.
Marien, P., Mampaey, E., Vervaet, A., Saerens, J., & DeDeyn, P. P. (1998). Normative data for the
Boston Naming Test in native Dutch-speaking Belgian elderly. Brain and Language, 65, 447–467.
Neils, J., Baris, J. M., Carter, C., Dell’aira, A. L., Nordloh, S. J., & Weiler, E. Q., et al. (1995). Effects of
age, education, and living environment on Boston Naming Test performance. Journal of Speech and
Hearing Research, 38, 1143–1149.
Nicholas, L. E., Brookshire, R. H., MacLennan, D. L., Schumacher, J. G., & Porrazzo, S. A. (1989).
Revised administration and scoring procedures for the Boston Naming Test and norms for non brain
damaged adults. Aphasiology, 3, 569–580.
Nicholas, M., Obler, L. K., Albert, M. L., & Goodglass, H. (1985). Lexical retrieval in healthy aging.
Cortex, 21, 595–606.
Quinones-Ubeda, S., Pena-Casanova, J., Bohm, P., Gramunt-Fombuena, N., & Comas, L. (2004).
Preliminary normative data for the second edition of the Boston Naming Test for young Spanish
adults. Neurologia, 19, 248–253.
Randolph, C., Lansing, A., Ivnik, R., Cullum, C. M., & Hermann, B. P. (1999). Determinants of
confrontation naming performance. Archives of Clinical Neuropsychology, 14, 489–496.
Roberts, P. M., Garcia, L. J., Desrochers, A., & Hernandez, D. (2002). English performance of proficient
bilingual adults on the Boston Naming Test. Aphasiology, 16, 635–645.
Ross, T. P., & Lichtenberg, P. A. (1998). Expanded normative data for the Boston Naming Test for use
with urban, elderly medical patients. The Clinical Neuropsychologist, 12, 475–481.
Ross, T. P., Lichtenberg, P. A., & Christensen, B. K. (1995). Normative data on the Boston Naming Test
for elderly adults in a demographically diverse medical sample. The Clinical Neuropsychologist, 9,
321–325.
Rosselli, M., & Ardila, A. (1990). Normative data on Boston Diagnostic Aphasia Examination in a
Spanish speaking population. Journal of Clinical and Experimental Neuropsychology, 12, 313–322.
Schmitter-Edgecombe, M., Vesneski, M., & Jones, D. W. (2000). Aging and word finding: A comparison
of spontaneous and constrained naming tests. Archives of Clinical Neuropsychology, 15, 479–493.
Tallberg, I. M. (2005). The Boston Naming Test in Swedish: Normative data. Brain and Language, 94,
19–31.
Thompson, L., & Heaton, R. (1989). Comparison of different versions of the Boston Naming Test. The
Clinical Neuropsychologist, 3, 184–192.
Tombaugh, T. N., & Hubley, A. M. (1997). The 60-Item Boston Naming Test: Norms for cognitively
intact adults aged 25 to 88 years. Journal of Clinical and Experimental Neuropsychology, 19, 922–932.
Van Gorp, W., Satz, P., Klersch, M. E., & Henry, R. (1986). Normative data on the Boston Naming Test
for a group of normal older adults. Journal of Clinical and Experimental Neuropsychology, 8, 702–705.
Welch, L. W., Doineau, D., Johnson, S., & King, D. (1996). Educational and gender normative data for
the Boston Naming Test in a group of older adults. Brain and Language, 53, 260–266.
Williams, B. W., Mack, W., & Henderson, V. W. (1989). Boston Naming Test in Alzheimer’s disease.
Neuropsychologia, 27, 1073–1079.
Worrall, L. E., Yiu, E. M. L., Hickson, L. M. H., & Barnett, H. M. (1995). Normative data for the Boston
Naming Test for Australian elderly. Aphasiology, 9, 541–551.
THE BOSTON NAMING TEST IN GREEK
1169
Downloaded by [City University London] at 06:08 16 March 2012
APPENDIX A
Number of correct responses per item in descending order
No Item Greek name
Number
correct
Order
in BNT No Item Greek name
Number
correct
Order
in BNT
1 scissors Yalı´di 100 6 31 muzzle Wı´mvtro 74 44
2 comb Xte
´
na/
tsatsa
´
ra
100 7 32 compass Diabg
´
tgz 74 50
3 saw Prio´ni 100 9 33 mask Ma
´
ska 73 18
4 hangar Krema
´
stra 100 15 34 harmonica Qusarmo´ nika 73 30
5 wreath SteQa
´
ni 100 28 35 pelican Peleka
´
noz 72 41
6 bed Kreba
´
ti 99 1 36 noose Hglia
´
72 48
7 octopus Xtapo´di 99 13 37 globe Udro´ceioz
sQaı´ra/
udro´ceioz
70 27
8 tree De
´
ntro 98 2 38 escalator Kulio´menez
ska
´
lez
69 37
9 snail Salicka
´
ri 97 22 39 dart Be
´
loz 62 25
10 broom Skou
´
pa 96 12 40 harp A
´
rpa 62 38
11 wheelchair Anapgriko´
karo´tsi/
anapgriko´
amajı´dio/
karotsa
´
ki
16 16 41 tongs Labı´da/tsimpı´da 60 54
12 house Spı´ti 94 4 42 protractor Moirocnvmo´nio 58 59
13 whistle SQurı´xtra 94 5 43 bun Tsoure
´
ki 57 19
14 funnel Xvnı´ 94 46 44 tripod Trı´podaz 56 52
15 flower Loulou
´
di 93 8 45 stilts Julopo´dara 55 34
16 helicopter Eliko´ptero 93 11 46 seahorse Ippo´kampoz 54 24
17 camel Kamg
´
la 93 17 47 hammock Aiv
´
ra 54 39
18 pressure
gauge
Pieso´metro 92 42 48 asparagus Spara
´
cci 50 49
19 pencil Molu
´
bi 91 3 49 latch Ma
´
ntalo/
su
´
rtgz
46 51
20 mushroom Manita
´
ri 91 14 50 sphinx sQg
´
cka 46 55
21 toothbrush Odonto´ bourtsa 88 10 51 column Kı´onaz 43 53
22 accordion Akornteo´ n 87 47 52 igloo Icklou
´
39 33
23 bench Packa
´
ki 85 20 53 abacus arihmgtg
´
rio 38 60
24 racquet Rake
´
ta 85 21 54 palette Pale
´
ta 37 58
25 pyramid Puramı´da
80 43 55 canoe Kano´3526
26 rhinoceros Rino´keroz 79 31 56 dominoes Nto´mino 35 35
27 mailbox Crammatokibv
´
t-
io
77 40 57 beaver Ka
´
storaz 28 29
28 volcano GQaı´steio 76 23 58 yoke fuco´ z 27 56
29 acorn Belanı´di 74 32 59 trellis Pe
´
rckola 26 57
30 cactus Ka
´
ktoz 74 36 60 unicorn Mono´kervz 18 45
The total number of correct responses was 4287 and the mean percentage correct for all participants is
71.4.
1170 PATRICACOU ET AL.
Downloaded by [City University London] at 06:08 16 March 2012
