Regensburg Word Fluency Test

Abbreviation: RWT

Speed-oriented mind games have been known for centuries. People have long passed the time with playing Categories or City - Country - River.

Word fluency procedures were probably first introduced into psychological testing by Thurstone & Thurstone in the Primary Mental Abilities Test¹⁾. Instead of the written word fluency assessment used by the Thurstones (and, accordingly, later also in the German LPS²⁾ ), Benton described somewhat later³⁾ an oral test form, the Controlled Verbal Fluency Task (CVFT): Within 60 seconds each, as many words as possible beginning with the letters F, A, and S must be named. The “F-A-S Test” was later developed into the Controlled Oral Word Association Test (COWA)⁴⁾ with two parallel versions (letters C, F, and L, or P, R, and W, respectively). The letters used occur about equally often as initial letters of English words.

In addition to such lexical word fluency tasks, semantic ones were also constructed. In the USA, the Set Test⁵⁾ has become established, which requires the rapid enumeration of colors, animals, fruits, and cities. In the testing of persons with dementia (for example with the CERAD or RBANS⁶⁾), semantic word fluency tasks are also used, although usually only a single category is required.

In the German-speaking region, such procedures were used until about 2000 only ad hoc for research purposes and without formal norming. Since 2000, the RWT⁷⁾ has been available, a procedure that contains a series of lexical and semantic word fluency tests that were standardized together on a German normative sample.

The RWT consists of 14 subtests.

Five subtests measure formal-lexical word fluency with the initial letters S, P, M, K, and B. In the German vocabulary, words with the initial letter S are very frequent, those with P and M frequent, and those with K and B less frequent.
Two subtests measure formal-lexical category switching, a task in which words with two different initial letters must be named alternately. For this, the initial letters G alternating with R and H alternating with T are used.
Five further subtests assess semantic-categorical fluency with the categories first names, animals, foods, occupations, and hobbies. A similar frequency distribution applies here as for lexical fluency: For first names the search space is very large, for animals and foods large, and for occupations and hobbies small.
The last two subtests assess the speed of semantic category switching between sports and fruits and between clothing items and flowers. As with formal-lexical category switching, a medium-sized search space can be assumed here as well.

In TDB2Online, only five of the 14 subtests are included, two from the formal-lexical paradigm (the letters K and S) and one each from the other three fluency paradigms. The TDB2Online norms are particularly relevant when the RWT is used together with other test procedures and the test results are to be considered together. This is typical for initial examinations. In an initial examination, it would be quite possible and sensible to limit oneself to the five selected subtests. Table 1 shows the various subtests in context.

Table 1: Subtests of the RWT

Fluency paradigm	Subtests	in TDB2Online
formal-lexical word fluency	S-word test	no
	P-word test	no
	M-word test	no
	K-word test	yes
	B-word test	no
formal-lexical category switching	Switching G-words/R-words	yes
	Switching H-words/T-words	no
semantic-categorical fluency	First names	no
	Animals	no, but see Discussion
	Foods	yes
	Occupations	no
	Hobbies	no
semantic category switching	Sports-fruits	no
	Clothing items-flowers	yes

The Animals subtest is discussed in the discussion section of this documentation because the largest number of independent norming studies exists for this subtest and these can quite well be compared across different languages. This does not apply to the lexical subtests, because their initial letters span search spaces of different sizes in different languages. The Animals subtest can therefore be used to roughly assess how the norming of the RWT fits into the international context.

The RWT is a copyright-protected test. The test material, including the test manual, can be obtained through the Testzentrale (test distributor) in Germany or Switzerland. Once one has become familiar with the administration using the manual, a single test administration requires only the separately available record forms and a (stop)watch. For each fluency paradigm there is one type of record form on which the corresponding subtest can be checked.

The normal test duration in the RWT is two minutes per subtest. The TDB2Online norms also apply to a test duration of two minutes. To nevertheless ensure administration of the test with examinees/patients of limited stamina, a norming for the first minute in each case was additionally carried out.

Internationally, a test duration of one minute is more common. The comparison data compiled in the discussion of the animal paradigm are therefore based only on one-minute measurements.

For each subtest, the raw score is the number of correctly produced words per test time (one or two minutes). The RWT instructions for each subtest are quite precise. It is always explained in detail what is counted as a correct response.

Because all norms of the RWT are (exclusively) available as percentile rank norms, the distributions of the normative sample could easily be back-calculated. For this, the overall norms broken down by age were used. Figures 1 to 5 show these back-calculated distributions for the subtests K words, S words, G-R words, Foods, and Clothing-Flowers.

Figure 1: Raw score distributions in the subtest K words for the five age groups of the RWT normative sample

Figure 2: Raw score distributions in the subtest S words for the five age groups of the RWT normative sample

Figure 3: Raw score distributions in the subtest G-R words for the five age groups of the RWT normative sample

Figure 4: Raw score distributions in the subtest Foods for the five age groups of the RWT normative sample

Figure 5: Raw score distributions in the subtest Clothing-Flowers for the five age groups of the RWT normative sample

These figures first of all show that the distributions are largely symmetrical and normal. With such verbal production tests, this is usually the case if they last long enough. The raw scores then lie in a range in which floor effects no longer occur. It is therefore quite astonishing that the authors of the RWT write in the manual, in a subordinate clause, that “the raw score distribution does not correspond to a normal distribution⁸⁾”. With many other tests, one would indeed wish that test authors were this critical in assessing the normality of their raw scores. Here, however, it is not necessary: at most a certain floor effect can be noted in the oldest group, most recognizable in subtest K. Otherwise, the raw scores of the RWT are as linear as one would like them to be for all tests.

The figures also show very clearly, on the other hand, that there are many bumpy curve courses both within each age group and between the age groups. This is the case with uncorrected percentile rank norms when the sample size is not 4 or 5 digits. Especially in the marginal areas, truncated distributions regularly occur with the usual sample sizes of a few hundred to a few thousand. One can see this quite well in the youngest age group in the subtest K words. Here the distribution breaks off abruptly at a raw score of 13, because this sample simply contained no examinee with lower values. For the 42-53-year-olds it is different: there was a particularly low-performing examinee with a raw score of 5, which, however, then causes the distribution shape to be unusual in this group as well. The matter can be made clear by converting the percentile rank distribution directly into standard scores (100;15). The result is shown in Figure 6.

Figure 6: Percentile-to-normal-transformed IQ-scaled standard scores of the subtest K words for the five age groups of the RWT normative sample

One can see that the two abnormal courses in the distribution curve appear here as deviations from the straight line (marked by arrows). Otherwise, the conversion of raw scores into standard scores is indeed very linear in all age groups.

Such sampling errors in the percentile rank norms of the RWT really must be corrected, because in individual cases they can have quite undesirable consequences. One could correct them heuristically by replacing only the ends of the distributions that deviate from linearity with linear extrapolations. As an example, we did this for the youngest age group in the subtest K words. Figures 7 and 8 show how a linear extrapolation (shown here as a dashed line) affects the distribution shape.

Figure 7: Percentile-to-normal-transformed IQ-scaled standard scores of the subtest K words for the five age groups of the RWT normative sample with linear extrapolation in the marginal areas (dashed)

Figure 8: Raw score distributions in the subtest K words for the five age groups of the RWT normative sample with linear extrapolation in the marginal areas (dashed)

With this method, the grossest errors at the margins could be corrected sensibly, but the various discontinuities in the curve course would remain. In addition, the procedure is somewhat laborious. Since the raw scores are very linear in all four subtests anyway, we ultimately decided on a simple linear transformation of the raw scores into standard scores.

The RWT manual reports neither means nor standard deviations of the raw scores. To estimate them, the standard scores calculated from the percentile ranks (as in Figure 7) were replaced by their linear trend, the latter being estimated only from raw scores corresponding to percentile ranks between 10 and 90. This procedure prevents a single, very abnormal value at the end of a distribution from decisively influencing the trend. This yielded the conversion of raw scores into standard scores, as shown in Figure 9 by way of example for the subtest K words. The means and standard deviations of the raw scores were estimated from the trend lines. The theoretical raw score distributions derived from them are shown in Figure 10.

Figure 9: Linear IQ-normed standard scores of the subtest K words for the five age groups of the RWT normative sample

Figure 10: Theoretical raw score distributions in the subtest K words for the five age groups of the RWT normative sample

In the subtest G-R words, one age group (the 42-53-year-olds) showed a considerably smaller spread than the other groups. This was already seen in Figure 3. This too must be regarded as a sampling error, because there is no sensible reason why a middle age group should have a markedly smaller standard deviation than the two neighboring age groups. Such sampling errors cannot be corrected by linearization within a single age group alone. This requires adjustments across the age groups. There are various heuristic (for example⁹⁾ ) and parametric procedures (e.g., inferential norming, which was used for the WISC-IV and WAIS-IV¹⁰⁾¹¹⁾ ), which, however, are somewhat laborious. We limited ourselves to adjusting the means and standard deviations by a third-order moving average, that is, with the respectively neighboring age groups, which could be carried out relatively easily and quickly. The marginal groups 18-29 and over 65 remained unchanged.

The test manual of the RWT presents norms that - depending on the subtest - were collected from between 532 and 634 adults. The manual says nothing about the targeted sample. In the manual, the sample is broken down by age (relatively coarsely into five age groups: 18-29, 30-41, 42-53, 54-65, and over 65 years), sex, and years of schooling (under 10 years of schooling, 10 to 12, and over 12 years). The norms were calculated across the total sample. They are given in the manual as pure percentile rank norms.

We are not aware of further norming studies in the German-speaking area for the initial letters and measurement times used in the RWT. For the domain of lexical word fluency, norming studies from other language areas cannot be used, because the number of words with a particular initial letter differs from language to language (more on this in the Discussion).

The situation is somewhat different for semantic word fluency. Here, at least for the category “animals”, there are many norming studies that can be compared - even across languages - with the corresponding subtest of the RWT. We carried out meta-analytic calculations for this, which are presented in the Discussion. They provide good information on how the RWT norms compare with other norms for this particular subtest.

In the absence of further data from other sources, the norms calculated for TDB2Online are based exclusively on the data given in the RWT manual. As described in the previous section, they are available in TDB2Online in linearized and smoothed form, which eliminated shortcomings of the original norms.

Figures 11 to 15 summarize the performance scores and the age norm limits of the five subtests as they appear in TDB2Online.

Figure 11 shows, in a composite graphic, the relationship between raw scores and performance scores as well as the (average) influence of age on performance in the subtest K words of the RWT. In this and all following figures, the vertical marks indicate the performance scores from 40 to 145. The raw scores are plotted in their correct positions on the black horizontal line. The alternately red and blue lines contain the norm limits for all age groups in the overview. The five markers on each line represent the percentile ranks 2.5, 16, 50, 84, and 97.5. One such line in each case, namely the one corresponding to the examinee's age group, is drawn (in a different form) in the TDB2Online profile sheet to aid in the individual interpretation of the test results. Because the performance scores are calculated from the results of the 18-29-year-old young adults, the 5 markers of the 18-29 age group lie exactly on the performance scores 70, 85, 100, 115, and 130.

Figure 11: Performance scores in the subtest K words with raw scores and age norm limits (see text)

In the first three decades of adulthood, no decline in performance can be seen in the subtest K words according to the RWT norms. A noteworthy decline can be seen only from age 50, which, however, even around age 70 amounts to only one standard deviation compared with young adults. Compared with tests that place high demands on memory or visuomotor function, this is rather little.

Figure 12 shows the relationship between raw scores and performance scores as well as the (average) influence of age on performance in the subtest S words of the RWT.

Figure 12: Performance scores in the subtest S words with raw scores and age norm limits (see text)

Although the search space for S words in German is about twice as large as for K words, the mean raw scores in the standardization sample lie very close together. The dependence of the norms on the search space is discussed in more detail in the Discussion.

Figure 13 shows the relationship between raw scores and performance scores as well as the (average) influence of age on performance in the subtest G-R words of the RWT.

Figure 13: Performance scores in the subtest G-R words with raw scores and age norm limits (see text)

It is astonishing how well the younger examinees cope, on average, with the additional load of alternating the initial letter. The average output over 2 minutes in the subtests K words and G-R words is practically level at around 25 words in 2 minutes. Only at higher ages is a larger decline in performance found in the alternating task. Losses in flexibility presumably have an effect here.

Figure 14 shows the relationship between raw scores and performance scores as well as the (average) influence of age on performance in the subtest Foods of the RWT.

Figure 14: Performance scores in the subtest Foods with raw scores and age norm limits (see text)

Rapidly naming foods is much easier than the two lexical tasks; here the mean of the young adults is around 40 words in two minutes.

Figure 15 shows the relationship between raw scores and performance scores as well as the (average) influence of age on performance in the subtest Clothing-Flowers of the RWT.

Figure 15: Performance scores in the subtest Clothing-Flowers with raw scores and age norm limits (see text)

In the subtest Clothing-Flowers, a somewhat different age course can be found than in the subtest Foods. Here the highest performance is reached only in later age groups. Possibly developments of interests are reflected here. Young people have presumably gathered less experience with both items of clothing and flowers than older people.

The psychometric processing of the RWT for graphical presentation in TDB2Online went largely without problems; a simple linear transformation served its purpose for all subtests, even though the required sample statistics first had to be derived from the percentile rank norms. The status of the procedure as a test for a particular aspect of executive functions is also clearly defined and well discussed in the neuropsychological literature worldwide. The RWT thus deserves its place in the spectrum of TDB2Online procedures, even though, in view of the narrow focus of word fluency tests, neuropsychological interpretations of RWT results are more limited in content than is the case for some other tests, and it is therefore probably also used less often.

The main problem with this German version of word fluency tests is - once again - the lack of representativeness of the norming. Hardly any indications can be drawn from the description of the sample; representativeness is not even claimed there. Independent data of healthy examinees from the German-speaking region on the RWT subtests could not be found until the end of 2012. In the former German-language Yahoo group Neuropsychologie, several clinical users noted in February 2004 that the test is “too difficult” and that even healthy and normally performing persons obtain scores in the below average range.

For the tests included in TDB2Online, we generally try to broaden the normative data situation and also draw on international sources for this when this is justified in terms of content. For the RWT subtests, this is unfortunately possible only very indirectly. For the formal-lexical subtests, results in other languages cannot be used at all, because the frequency of words with particular initial letters differs between languages. We did, however, try to reconcile the results of two other German-language publications with the results of the RWT. This attempt is presented further below. For the categorical-semantic subtests, by contrast, a comparison across language boundaries is quite possible. At least for the category Animals, there are sufficient international data to allow a comparison, although only for the measurement time of one minute. This is also presented in the following.

In the German-speaking region, there is another test that measures formal-lexical word fluency, the LPS. It exists in two versions, the original LPS of 1962¹²⁾ and the LPS+ version with normative data for older persons from 1993¹³⁾. In the new LPS-2 of 2013¹⁴⁾, the word fluency task is no longer included. The task in the LPS and LPS 50+ is, however, different from that in the RWT: Here, as many words as possible beginning with the letters L, P, and R (parallel version: F, K, and R) must be written for 1 minute each.

In the LPS, then, other letters serve as initial letters for the words, and in addition one must write them instead of naming them. The latter is likely to have effects in the upper performance range in any case, but not necessarily in the lower or middle range. There is a certain relationship between the frequency of words with a particular initial letter in everyday language and the association frequency for words with this initial letter, even if it is not overly close. In Table 2, the mean one-minute association raw scores of adults for initial letters that occur in German word fluency tests are compared with the percentage frequencies of the same initial letters in the general-language German dictionary elexicon on the OWID portal of the Institute for the German Language in Mannheim. The table also lists the information on the search space of the respective initial letters taken from the RWT manual.

It can be seen that examinees think of fewer words for the rarely occurring initial letters M and P than for normally frequent ones such as B or K. Upward, by contrast, the gain appears to be limited: even with the very frequently occurring letter S, one thinks of hardly more words in one or even two minutes than with the normally frequent letters B and K. While the elexicon data and the one-minute norms of the RWT (here for 18-29-year-olds) agree well, the verbal search space statements in the RWT manual deviate from them for most letters. Only for the letter S do the statements agree. The RWT manual does not mention where the search space statements come from. It makes a big difference whether dictionary entries (which is probably the right reference set) or running texts are analyzed for this. In running texts, for example, the letter D occurs much more frequently than in lexica because of the many definite articles that, in German, all start with D.

Table 2: Number of words in general-language German dictionaries and mean word fluency raw scores as a function of initial letter

Letter	Word frequency	Search space	mean raw scores of word fluency tests
			Duration in s	RWT	LPS	LPS 50+
S	12.0	very large	60	16
B	6.8	small	60	15
K	7.0	small	60	15.5
M	5.1	large	60	13.7
P	4.9	large	60	12
L+P+R	3.4 + 4.9 + 4.4		60 each		26	31
F+K+R	5.5 + 7.0 + 4.0		60 each		26	31

_{Word frequency: Percentage frequency of words with this initial letter in the general-language dictionary elexicon}
_{Search space: Statement in the RWT manual}
_{Norming date RWT: 2000, LPS: 1962, LPS 50+: 1993}
_{Age group RWT: 18-29 years, LPS: 18-29 years, LPS 50+: 50-69 years}

The table also includes the norms of Subtest 6 of the Leistungsprüfsystem LPS (Horn, 1962) and the LPS 50+ (Sturm, Willmes & Horn, 1993). In both subtests, as many words as possible are to be written for three different letters over one minute each, and the norms are available only for the sum of the three minutes. It can nevertheless be seen that the LPS normative sample shows much lower performance than that of the RWT. For the three letters given in the LPS, it would be expected that the young examinees of the RWT sample would probably have produced 30 to 33 words rather than only 26, which is given as the norm in the LPS manual. However, it can also be seen that the 50-69-year-olds of the LPS 50+ normative sample (this is the youngest age group for this version of the LPS), with 31 words, performed considerably better than the 18-29-year-olds of the LPS normative sample (26 words). The performance of the “over 50-year-olds” in the LPS normative sample was in fact only 18 words.

Possibly two effects add up here: On the one hand, given the large time gap between the three normings (LPS 1962, LPS50+ 1993, RWT 2000), cohort effects in the sense of an “IQ gain” must be assumed. Accordingly, later normings must be stricter, even though there are no data on the extent of the IQ gain in the domain of word fluency. On the other hand, there also seem to have been different demands regarding the representativeness of the norms. Only in the old LPS norming can one read that the author made an effort to systematically seek out the weaker students (e.g., those who had repeated a grade) of a cohort. The other two normings are rather “incidental” normative samples, in which more capable and more motivated examinees are generally overrepresented.

Ultimately, the comparison remains of little informative value. Compared with the LPS, the RWT norms would be too strict; compared with the LPS 50+, they would be comparable. In both comparisons there are further limitations, which is why no real conclusions can be drawn.

For the Animals subtest, a relatively large amount of data exists internationally, although only for a duration of one minute. Language does not play a major role for performance on this task, as has been demonstrated in various studies¹⁵⁾ ¹⁶⁾ ¹⁷⁾. In the second edition of their handbook, Mitrushina et al.¹⁸⁾ compiled 45 studies with results of word fluency tests in healthy examinees. Twenty of these reported data on the “animal” paradigm for examinee groups that could be delimited by age. The results of these 20 studies, together with the data from the RWT manual and the Basel normative data for the CERAD, were analyzed in a meta-analysis of age-dependent performance in the one-minute range. Table 3 gives an overview of the included studies, in which a total of 8488 persons were examined. Most studies come from the USA. The educational level of the participants was extremely varied and ranges (on average!) between 6.7 and 16.5 years.

Table 3: Overview of international data on the fluency paradigm “Animals”. Duration 1 minute.

Sample	Year	N	Education (years)	Country
Selnes¹⁹⁾	1991	696	16.5	USA
Kozora²⁰⁾	1995	174	14.5	USA
Crossley²¹⁾	1997	635	10	Canada
Beatty²²⁾	1997	38	13.4	USA
Salthouse²³⁾	1997	115	15.3	USA
Kempler²⁴⁾	1998	317	10.3	USA
Stuss²⁵⁾	1998	62	13.9	USA
Johnson-Selfridge²⁶⁾	1998	600	13.2	USA
Tombaugh²⁷⁾	1999	735	11.4	USA
Epker²⁸⁾	1999	65	14.3	USA
Gladsjo²⁹⁾	1999	768	13.6	USA
Binder³⁰⁾	1999	125	13.5	USA
RWT³¹⁾	2000	634	?	D
Fama³²⁾	2000	51	16.4	USA
Acevedo³³⁾	2000	553	14	USA
Brady³⁴⁾	2001	235	14	USA
CERAD³⁵⁾	2002	1100	12.5	CH, D, A
Grady³⁶⁾	2002	1063	12.7	USA
Giovannetti³⁷⁾	2003	31	15	USA
Lopez-Carlos³⁸⁾	2003	115	6.7	USA, Mexico
Miller³⁹⁾	2003	227	16.3	USA
Ravdin⁴⁰⁾	2003	149	15.6	USA

In Figure 16, the results of the 22 studies from Table 3 are presented graphically. In the figure, most studies are plotted as thin lines. Only 6 curves are marked in bold: in black, the raw scores averaged across all studies and their 1-sigma limits, in blue the norms of the RWT, and in yellow and magenta the curves of the two CERAD subsamples (up to and over 12 years of schooling).

Figure 16: Age curves of the one-minute raw scores for “Animals” of the studies from Table 3 and their meta-analytic summary

It is first apparent that the influence of age on semantic word fluency is not as large as for some other functions. The difference between the mean of the young adults and that of the 80-year-olds is about one standard deviation. For visuomotor functions, it is regularly closer to two standard deviations. Despite the very heterogeneous samples, most of the data of the individual studies also lie within the one-sigma limits. The Basel CERAD sample proves to be extremely high-performing. Here, the subsample with more than 12 years of schooling lies above the one-sigma limit, and the sample with up to 12 years of schooling also still lies clearly above most other studies in this age range. The RWT sample cannot be evaluated across the board in this paradigm: The young examinees perform comparatively very well and those over 65 comparatively very poorly when compared with the means of all samples. This probably reflects selective sampling, in which the young subsamples have a much higher level of education than the older ones. In part this must be so because of the steadily increasing education of the population, but possibly the effect is very pronounced in the RWT sample.

The results of this analysis can presumably be extrapolated to the other RWT subtests. The samples do not differ noticeably; the N varies between 532 and 634 across the individual subtests, probably unrelated to the sample characteristics that are responsible for the particular curves of the RWT in Figure 16. Accordingly, the RWT norms are probably about half a standard deviation too strict for the 18-39-year-olds. For the 40-60-year-olds, this difference progressively decreases, and for those over 65, the norms are then about half a standard deviation too lenient. Since the TDB2Online norms for all two-minute subtests are based exclusively on the RWT norms, for lack of suitable data from other sources, this evaluation naturally also applies to the values calculated in TDB2Online.

¹⁾

Thurstone, L.L., & Thurstone, T.G.(1962). Primary mental abilities (Rev.). Chicago: Science Research Associates.

²⁾

Horn, W. (1983) L-P-S Leistungsprüfsystem. 2. Auflage. Göttingen: Hogrefe.

³⁾

Benton, A.(1967). Problems of test construction in the field of aphasia. Cortex, 3, 32-58.

⁴⁾

Benton, A., & Hamsher, K.(1978). Multilingual Aphasia Examination manual. Iowa City: Unversity of Iowa.

⁵⁾

Isaacs, B. & Kennie, A. T. (1973) The Set Test as an aid to the detection of dementia in old people. British Journal of Psychiatry, 123, 467-470.

⁶⁾

Randolph, C. (1998). Repeatable Battery for the Assessment of Neuropsychological Status (RBANS). San Antonio/Texas: Psychological Corporation

⁷⁾ , ³¹⁾

Aschenbrenner, A., Tucha, O. & Lange K. (2000) RWT. Regensburger Wortflüssigkeits-Test. Handanweisung. Göttingen: Hogrefe

⁸⁾

Aschenbrenner, A., Tucha, O. & Lange K. (2000) RWT. Regensburger Wortflüssigkeits-Test. Handanweisung. Göttingen: Hogrefe, S. 23 unten

⁹⁾

Pauker, J.D. (1988) Constructing overlapping cell tables to maximize the clinical usefulness of normative test data: Rationale and an example from neuropsychology. Journal of Clinical Psychology, 44, 930-933.

¹⁰⁾

Wilkins, C., Rolfhus, E., Weiss, L., & Zhu, J. J. (2005) A new method for calibrating translated tests with small sample sizes. Paper presented at the 2005 annual meeting of the American Educational Research Association, Montreal, Quebec, Canada

¹¹⁾

Zhu, J. J. & Chen, H.-Y. (2011) Utility of inferential norming with smaller sample sizes. Journal of Psychoeducational Assessment, 29, 570-580

¹²⁾

Horn, W. (1962) Leistungsprüfsystem (LPS). Göttingen: Hogrefe

¹³⁾

Sturm, K., Willmes, K. & Horn, W. (1993) Leistungsprüfsystem für 50-90jährige (LPS 50+). Göttingen: Hogrefe

¹⁴⁾

Kreuzpointner, L., Lukesch, H. & Horn, W. (2013) Leistungsprüfsystem 2 (LPS-2). Göttingen: Hogrefe

¹⁵⁾ , ²⁴⁾

Kempler, D., Teng, E. L. Dick, M., Taussig, I.M. & Davis, D. S. (1998) The effects of age, education, and ethnicity on verbal fluency. Journal of the International Neuropsychological Society, 4, 531-538.

¹⁶⁾ , ²⁶⁾

Johnson-Selfridge, M., Zalewski, C. & Aboudarham, J. (1998) The relationship between ethnicity and word fluency. Archives of Clinical Neuropsychology, 13, 319-325.

¹⁷⁾ , ³³⁾

Acevedo, A., Loewenstein, D. A., Barker, W. W., Harwood, D. G., Luis, C., Bravo, M. et al. (2000) Category Fluency Test: Normative data for English- and Spanish-speaking elderly. Journal of the International Neuropsychological Society, 6, 760-769.

¹⁸⁾

Mitrushina, M., Boone, K., Razani, J., D'Elia, L. (2005) Handbook of Normative Data for Neuropsychological Assessment, 2. Auflage. New York: Oxford University Press

¹⁹⁾

Selnes, O.A., Jacobson, L., Machado, A.M., Becker, J.T., Wesch, J., Miller, E.N., et al. (1991) Normative data for a brief neuropsychological screening battery. Perceptual and Motor Skills, 73, 539-550.

²⁰⁾

Kozora, E. & Cullum, C.M. (1995) Generative naming in normal aging: Total output and qualitative changes using phonemic and semantic constraints. Clinical Neuropsychologist, 9, 313-320.

²¹⁾

Crossley, M., D'Arcy, C. & Rawson, N.S.B. (1997) Letter and category fluency in community-dwelling Canadian seniors: A comparison of normal participants to those with dementa of the Alzheimer or vascular type. Journal of Clinical and Experimental Neuropsychology, 19, 52-62.

²²⁾

Beatty, W.W., Testa, J.A., English, S. & Winn, P. (1997) Influences of clustering and switching on the verbal fluency performance of patients with Alzheimer's disease. Aging, Neuropsychology, and Cognition, 4, 273-279.

²³⁾

Salthouse, T.A., Toth, J., Hancock, H.E. & Woodward, J.L. (1997) Controlled and automatic forms of memory and attention: Process purity and the uniqueness of age-related influences. Journals of Gerontology. Series B: Psychological Sciences and Social Sciences, 52B, P216-P228.

²⁵⁾

Stuss, D.T., Alexander, M.P., Hamer, L., Palumbo, C., Dempster, R., Binns, M., et al. (1998) The effects of focal anterior and posterior brain lesions on verbal fluency. Journal of the International Neuropsychological Society, 4, 263-278.

²⁷⁾

Tombaugh, T., Kozak, J. & Rees, L. (1999) Normative data stratified by age and education for two measures of verbal fluency: FAS and animal naming. Archives of Clinical Neuropsychology, 14, 167-177.

²⁸⁾

Epker, M.O., Lacritz, L.H. & Cullum, C.M. (1999) Comparative analysis of qualitative verbal fluency performance in normal elderly and demented populations. Journal of Clinical and Experimental Neuropsychology, 21, 425-434.

²⁹⁾

Gladsjo, J.A., Schuman, C.C., Evans, J.D., Peavy, G. M., Miller, S. W. & Heaton, R. K. (1999) Norms for letter and category fluency: Demographic corrections for age, education, and ethnicity. Assessment, 6, 147-178.

³⁰⁾

Binder, E.F., Storandt, M. & Birge, S. J. (1999) The relation between psychometric test performance and physical performance in older adults. Journals of Gerontology: Series A: Biological Sciences and Medical Sciences, 54, M428-M432.

³²⁾

Fama, R., Sullivan, E.V., Shear, P.K., Cahn-Weiner, D.A., Marsh, L., Lim, K.O. et al. (2000) Structural brain correlates of verbal and nonverbal fluency measures in Alzheimer's disease. Neuropsychology, 14, 29-41.

³⁴⁾

Brady, C. B., Spiro, A., III, McGlinchey-Berroth, R., Milberg, W. & Gaziano, J. M. (2001) Stroke risk predicts verbal fluency decline in healthy older men: Evidence from the normative aging study. Journals of Gerontoloty: Series B: Psychological Sciences and Social Sciences, 56B, P340-P346.

³⁵⁾

Link auf die Schweizer CERAD-Homepage

³⁶⁾

Grady, D., Yaffe, K., Kristof, M., Lin, F., Richards, C., & Barrett-Connor, E. (2002) Effect of postmenopausal hormone therapy on cognitive function: The Heart and Estrogen/Progestin Replacement Study. American Journal of Medicine, 113, 543-548.

³⁷⁾

Giovannetti, T., Goldstein, R. Z., Schullery, M., Barr, W.B. & Bilder, R. M. (2003) Category fluency in first-episode schizophrenia. Journal of the International Neuropsychological Society, 9, 384-393.

³⁸⁾

Lopez-Carlos, E., Salazar, X.F., Villasenor, T., Saucedo, C., & Pena, R. (2003) Validez y datos normativos de la pruebas de nominacion en personas con educacion limitada. Poster presented at the Neuropsicologia-Congreso Latinoamericano por la Sociedad Latinoamericana de Neuropsicologia, Toronto, Canada.

³⁹⁾

Miller, E.N. Personal Communication 2003, Data from Mitrushina et al., 2005

⁴⁰⁾

Ravdin, L.D., Katzen, H. L., Agrawal, P. & Relkin, N.R. (2003) Letter and semantic fluency in older adults: Effects of mild depressive symptoms and age-stratified normative data. Clinical Neuropsychologist, 17, 195-202.