Вы находитесь на странице: 1из 7

Brain Injury, 20 August 2005; 19(9): 685–691


Brief assessment of severe language impairments: Initial validation

of the Mississippi aphasia screening test


Methodist Rehabilitation Center, Jackson, MI, USA and 2University of Mississippi Medical Center, MI, USA

(Received 2 April 2004; accepted 8 November 2004)

Primary objective: To validate the Mississippi Aphasia Screening Test (MAST) which includes nine sub-scales measuring
expressive and receptive language abilities.
Research design: Evaluation of inpatients admitted to neurology, neurosurgery or rehabilitation units at two local hospitals and
who were within 60 days of onset of a unilateral ischemic or haemorrhagic stroke (left hemisphere (LH; n ¼ 38); right hemi-
sphere (RH; n ¼ 20)). Additional participants were recruited from the community to comprise a non-patient control sample
(NP; n ¼ 36).
Methods: Data collection included administration of the MAST and chart review.
Results: The LH group showed more impairment than the RH and NP groups on summary scores. The LH group performed
worse than the NP group on all sub-scales. The object recognition and verbal fluency sub-scales did not discriminate the
stroke groups.
Conclusion: Analyses suggest good criterion validity for the MAST in differentiating communication impairments among
clinical and control samples.

Keywords: Stroke, aphasia, assessment, screening, cognitive

batteries may not be useful for serial evaluation.

A brief instrument that surveys a broad array
Individuals with acquired brain injury may demon- of language and communication abilities is more
strate communication impairment, particularly suited to tracking patients’ progress through their
during early recovery. Focal left hemisphere lesions early clinical course to provide accurate information
may cause marked language impairment acutely and to both family and clinicians.
result in persistent aphasia with subsequent adverse Individuals with severe language impairment are
implications for treatment and long-term outcome challenging to evaluate. The length of comprehensive
[1–3]. The degree of language impairment among language batteries such as the Western Aphasia
severely aphasic individuals is often under-estimated Battery and Boston Diagnostic Aphasia Examina-
by doctors, nurses and relatives of patients [4]. tion require 90 minutes and up to 2 hours to admin-
Clinicians often serially assess communication skills ister (and may take up to 8 hours in some patients)
and language function. This yields information to [5–7]. Other measures have been developed that
guide medical and rehabilitation interventions, are relatively more brief (i.e. shortened version of
detect improvement or decline in clinical status, the Minnesota Aphasia Test, Boston Assessment
provide feedback to family and assist with post- of Severe Aphasia; and Multi-lingual Aphasia Exami-
discharge planning. For patients with severe nation) but still require 30–45 minutes in administra-
language impairments, lengthy language assessment tion time [8–12]. The major disadvantages of using

Correspondence: Risa Nakase-Thompson, PhD, Department of Neuropsychology, Methodist Rehabilitation Center, 1350 East Woodrow Wilson Drive,
Jackson, MI 39216, USA. Tel: 1.601.364.3448. Fax: 1.601.364.3558. E-mail: nakase@aol.com
ISSN 0269–9052 print/ISSN 1362–301X online # 2005 Taylor & Francis Group Ltd
DOI: 10.1080/02699050400025331
686 R. Nakase-Thompson et al.

these language batteries with severely aphasic hospitals (Level 1 trauma centre and rehabilitation
patients are that small improvements over time may hospital) and who received the MAST as part of
not be reflected in summary scores and patients neuropsychological consultations administered by
may be frustrated by the lengthy testing interval. neuropsychologists or specially trained psychometri-
Other language tests are brief in administration cians between 15 June 2001–31 December 2002 were
but evaluate a limited number of abilities such as potential subjects. All were within 60 days of the onset
the Boston Naming Test, Token Test or Controlled of a unilateral ischemic or haemorrhagic stroke.
Oral Word Association Test [11, 13, 14]. As a Individuals with unilateral strokes who were greater
result, common practice involves administering test- than 60 days post-stroke (n ¼ 8) and individuals with
ing batteries that encompass many different bilateral strokes (n ¼ 10) were excluded. Subjects who
individual tests across multiple sessions to assess did not receive a MAST as part of a bedside
patients’ evolving abilities [5, 15]. Performance examination or did not complete the entire MAST
differences from one part of the battery to another were not included. Thus, the study sample was not
may reflect evolution of symptoms between two consecutive. If subjects received more than one
testing sessions. The need for a brief screening MAST, only the first observation (one closest to the
instrument that could be completed within a single time of onset) was analysed for study purposes.
therapy session, bedside evaluation or clinic appoint- Participants were classified into either a left-
ment has been previously highlighted and develop- hemisphere (LH) or right-hemisphere stroke group
ment of such a measure would address some of the (RH) based on neuroimaging data.
limitations of comprehensive language batteries Non-patient control participant data were col-
[15, 17]. Some brief tests have been developed lected by occupational therapy students who were
to address some of these issues. For example, enrolled in an allied health profession programme.
the Frenchay Aphasia Screening Test (FAST) was These students were enrolled in a class focusing on
developed as a tool for non-cognitive specialists to administration of cognitive and functional clinical
identify persons with linguistic difficulties to be measures. They were trained in administration of
referred for more comprehensive assessment [12]. the MAST and were asked to collect data from
The FAST has four sub-scales (comprehension, friends and family. The students were instructed
expression, reading and writing) that identify to only collect data from individuals who were
probable aphasia. However, the FAST was not devel- 18 years of age or older and did not have a history
oped to assist with monitoring progress over time or of neurological dysfunction, such as stroke, trau-
goal planning [12]. The Aachen Aphasia Bedside matic brain injury or dementia. Participants recrui-
Test was developed in Germany as a measure of ted from the community comprised the non-patient
language functioning to be used with persons who control group (NP). These subjects were not
are in the first 1–2 months post-stroke but studies expected to demonstrate difficulties with the brief
of this measure are all in German and not available screening instrument.
in English [17].
The Mississippi Aphasia Screening Test (MAST) Measure
was developed as a brief, repeatable screening mea-
sure for individuals with severely impaired communi- The Mississippi Aphasia Screening Test (MAST) was
cation/language skills. Such a brief measure may be initially developed by a team of neuropsychologists,
advantageous for individuals with severe language physicians and speech-language pathologists. Fifteen-
impairments who may be frustrated and stressed by to-thirty items were initially generated for each sub-
lengthy testing sessions. The MAST was designed scale then pilot-tested and revised to form the current
for serial assessment of changes in language abilities test (see sample items in appendix 1). The current
over time and requires 5–10 minutes to administer. form has nine sub-tests: (a) Naming, (b) Automatic
The purpose of this paper is to demonstrate initial Speech, (c) Repetition, (d) Yes/No Accuracy, (e)
psychometric and clinical properties of the MAST Object Recognition from a Field of Five, (f ) Following
and to provide descriptive information regarding Verbal Instructions, (g) Reading Instructions, (h)
performance of patient and non-patient groups on Verbal Fluency, and (i) Writing/Spelling to Dictation.
this test. The test takes 5–10 minutes to administer and yields
nine sub-test scores, two index scores (Receptive &
Expressive Language Index) and a total score. Each
Methods sub-scale sums to 10 points with the exception of
the Yes/No Accuracy sub-scale which sums to 20.
The Index scores sum to 50 points each and are
Patients with stroke admitted to neurology, neuro- added for the MAST Total Score (0–100 points).
surgery or rehabilitation units at two local Stimulus materials include one photograph, five
Severe language impairments 687

written instructions and five common objects using the MAST individual sub-tests, MAST
available in most hospital or clinic settings. Expressive and Receptive Indices and MAST Total
The MAST is available online or by contacting the Score, respectively.
first author [18].
Chart review
Thirty-eight individuals comprised the left-
Demographic and medical information was abstrac- hemisphere stroke group (LH) and 20 comprised
ted by chart review in both the Level I trauma centre the right-hemisphere stroke group (RH). Thirty-six
and rehabilitation hospital. Information abstracted subjects were recruited from the community to
included age, gender, handedness, educational level, comprise the non-patient control sample (NP). A
neuroimaging findings and duration of time since summary of demographic variables is presented by
stroke to administration of initial MAST. group in Table I. Educational level was unknown
for 12% of the participants and handedness was
unknown for 3% of the participants. ANOVAs
Statistical analyses
revealed significant between-group differences both
Descriptive statistics were generated by group for the for age, F(2, 90) ¼ 8.67, p < 0.0005 and educational
demographic variables of age, educational level, level, F(2, 79) ¼ 4.28, p ¼ 0.017. Post-hoc pairwise
handedness and gender as well as for the nine comparisons utilizing a liberal level of significance for
sub-scales and three summary scores of the MAST. interpretation (Bonferroni level of significance is
Time since stroke, which was the duration of time in 0.017 for three groups) indicated the NP group to be
days between the patient’s stroke and administration significantly younger than either the LH group
of the MAST was also calculated for the patient ( p < 0.0005) or the RH group ( p ¼ 0.025). The NP
groups. Analyses of variance (ANOVAs), indepen- group had significantly more years of education than
dent t-tests or chi-square analyses, as appropriate, the RH group ( p ¼ 0.032). No significant difference
were conducted to assess for between-group was found to exist between groups with respect to
differences across demographic variables. Pearson’s gender, 2(2, n ¼ 93) ¼ 3.10, p > 0.05, or handed-
product moment correlation coefficients were com- ness, 2(2, n ¼ 90) ¼ 0.94, p > 0.05. Additionally, the
puted to examine the degree of association between patient groups were not found to differ in duration of
demographic variables and MAST scores in the time between the stroke and administration of the
control group. In order to protect against inflation MAST, t(55) ¼ 0.69, p > 0.05.
of Type I error, multi-variate analysis of covariance Results of correlational analyses between MAST
(MANCOVA) was utilized to assess for between- performance and demographic variables in the
group differences across the 12 measures of the NP group are presented in Table II; p-values are
MAST, controlling for differences between groups reported for descriptive purposes only and are not
with respect to demographic variables that were found corrected for multiple comparisons. Correlations
to be significantly associated with test performance. were defined as significant for purposes of covar-
Significant between-group effects on the MAST were iance, however, only when p-values were <0.01.
examined using Bonferroni post-hoc analyses. Three Age, education and handedness were significantly
independent discriminant function analyses were also associated with scores on the Expressive Index
computed to determine classification rates by group and Total Score. Age and handedness were also

Table I. Demographic characteristics of sample by group.

Patient groups

Non-patient group LH RH p

Age, M (SD) 46.6 (19.2) 61.7 (12.7) 58.7 (15.7) <0.0005

Education, M (SD) 13.8 (2.1) 12.2 (3.6) 11.6 (2.3) 0.017
Duration, M (SD) — 22.2 (15.6) 19.3 (14.8) >0.50
Gender >0.05
Male 36% 49% 60%
Female 64% 51% 40%
Handedness >0.05
Left 11% 5% 5%
Right 89% 87% 95%
Unknown 0% 8% 0%

Note: Age and education are measured in years, duration in time is characterized in days from stroke to MAST
assessment, LH ¼ Left Hemisphere Stroke, RH ¼ Right Hemisphere Stroke.
688 R. Nakase-Thompson et al.

Table II. Correlations between demographic variables and MAST scores in non-patient group.

MAST scale Age Education Gender Handedness

Naming — — — —
Automatic speech 0.07 0.14 0.13 0.06
Repetition — — — —
Yes–no 0.25 0.15 0.32 0.09
Writing to dictation 0.21 0.23 0.13 0.06
Object recognition — — — —
Following verbal instructions 0.15 0.01 0.15 0.08
Following written instructions 0.12 0.21 0.07 0.09
Verbal fluency 0.36* 0.28 0.03 0.37*
Expressive index 0.39* 0.33* 0.08 0.37*
Receptive index 0.16 0.03 0.33* 0.15
Total score 0.44y 0.34* 0.01 0.41*

Note: MAST ¼ Mississippi Aphasia Screening Test.

*p < 0.05, y p < 0.01, Missing numbers indicate no variance among items correlated.

Table III. MAST performance by group.

Patient groups

MAST scale Non-patient group LH RH F ( p)* 2

Naming 10.0 (0) 4.3 (4.2) 7.6 (3.2) 25.0 (<0.0005)abc 0.36
Automatic speech 9.9 (0.7) 5.4 (3.8) 8.4 (3.1) 16.8 (<0.0005)ab 0.28
Repetition 10.0 (0) 4.8 (4.0) 8.2 (3.3) 23.2 (<0.0005)ab 0.34
Yes–no 19.9 (0.5) 14.5 (5.7) 17.8 (4.4) 11.2 (<0.0005)ab 0.20
Writing to dictation 9.9 (0.7) 1.7 (3.4) 5.5 (4.1) 55.6 (<0.0005)abc 0.56
Object recognition 10.0 (0) 6.2 (4.4) 8.1 (3.7) 8.7 (<0.0005)a 0.16
Following verbal instructions 9.9 (0.4) 4.5 (3.7) 7.8 (3.5) 25.2 (<0.0005)abc 0.36
Following written instructions 9.9 (0.5) 2.6 (3.7) 6.6 (3.7) 45.4 (<0.0005)abc 0.51
Verbal fluency 6.4 (3.5) 3.1 (4.3) 5.3 (4.4) 3.2 (<0.05)a 0.07
Expressive index 46.2 (3.7) 19.3 (16.2) 34.9 (15.8) 29.9 (<0.0005)abc 0.40
Receptive index 49.7 (0.7) 27.9 (15.1) 40.3 (14.2) 24.4 (<0.0005)abc 0.35
Total score 95.9 (3.6) 47.2 (29.9) 75.2 (29.1) 29.9 (<0.0005)abc 0.40

Note: MAST ¼ Mississippi Aphasia Screening Test; LH ¼ Left Hemisphere Stroke; RH ¼ Right Hemisphere Stroke.
Left hemisphere patient group scores significantly lower than non-patient group scores, b Left hemisphere patient group scores significantly
lower than right hemisphere patient group scores, c Right hemisphere patient group scores significantly lower than non-patient group scores.
* degrees of freedom ¼ (2, 89). Group performance scores are means and standard deviations.

significantly associated with scores on the Verbal Additionally, the RH group performed significantly
Fluency sub-test and gender was significantly poorer than did the NP group on seven of the 12
associated with scores on the Receptive Index. MAST scores ( ps < 0.03).
Mean scores of the groups across the nine sub- Results of discriminant function analyses yielded
scales and three summary scores of the MAST are correct classification rates of 74.2%, 69.9% and
presented in Table III. MANCOVA using age as a 71.0% for the sample using the MAST individual
covariate, indicated a significant main effect for sub-tests, MAST Expressive Index and Receptive
group, F(18, 162) ¼ 6.26, p < 0.0005. In a separate Index and MAST Total Score, respectively. As can
analysis, the main effect was found to remain signifi- be seen in Table IV, the majority of incorrect classi-
cant in the absence of age covariation. Subsequent fications occurred for the RH group (30% correct)
examination of univariate analyses revealed signifi- since the majority were misclassified as being in the
cant between-group differences across all 12 scores non-patient group. In all analyses, correct classifica-
of the MAST, with partial eta-squared effect sizes tion rates exceeded 62% for the LH group and 97%
ranging from 0.07–0.56 (see Table III). As expected, for the NP group.
results of post-hoc pairwise comparisons indicated
the LH group performed significantly poorer than
did the NP group across all 12 MAST scores
( ps < 0.003). The LH group also performed sig- Current findings indicate strong evidence for the
nificantly poorer than did the RH group on 10 criterion validity of the MAST. MAST total, index
of the 12 measures of the MAST ( ps < 0.02). and sub-scale scores differentiated individuals
Severe language impairments 689

Table IV. Group classification based on discriminant functional analysis.

Predicted group membership

Actual group membership Left hemisphere stroke Right hemisphere stroke Non-patients

MAST sub-tests
Left hemisphere stroke 28 (75.7%) 6 (16.2%) 3 (8.1%)
Right hemisphere stroke 6 (30.0%) 6 (30.0%) 8 (40.0%)
Non-patients 0 (0%) 1 (2.8%) 35 (97.2%)
Total correct ¼ 69/93 (74.2%)
MAST expressive index and receptive index 4 (10.8%)
Left hemisphere stroke 23 (62.2%) 10 (27.0%) 9 (45.0%)
Right hemisphere stroke 5 (25.0%) 6 (30.0%) 36 (100%)
Non-patients 0 (0%) 0 (0%)
Total correct ¼ 65/93 (69.9%)
MAST total score
Left hemisphere stroke 24 (64.9%) 9 (24.3%) 4 (10.8%)
Right hemisphere stroke 4 (20.0%) 6 (30.0%) 10 (50.0%)
Non-patients 0 (0%) 0 (0%) 36 (100%)
Total correct ¼ 66/93 (71.0%)

Note: MAST ¼ Mississippi Aphasia Screening Test.

with left hemisphere stroke from normal controls between the RH group and NP group resulting
and differentiated patients with left hemisphere in many RH individuals being misclassified as NP.
stroke from those with right hemisphere stroke. However, the RH group performed more poorly
MAST scores are sensitive to language impairments than the NP group on four of nine MAST sub-
associated with dominant hemisphere injury. scale scores as well as on the MAST total and
As expected, patients in the LH group scored index scores. These findings may indicate that
more poorly than the NP group on all MAST total, performance on the MAST is sensitive to cognitive
index and sub-test scores. The MAST demonstrates impairments associated with right hemisphere
excellent sensitivity and specificity in detecting stroke albeit to a lesser degree than by cognitive
language impairment and correctly discriminating impairments associated with left hemisphere stroke.
language impaired individuals (LH group) from As noted above, some of the sub-scales of the
normal controls (NP group). Using summary MAST have visual-perceptual demands. These sub-
scores, none of the NP group members were misclas- scales include object recognition from a field of five
sified yielding 100% correct classification for non- and following written instructions. These tasks can
patients. Depending on which sub-tests or summary be particularly difficult for patients with left neglect.
scores were used, 62–76% of the LH group This visual analysis component of the MAST may
members were correctly classified. As expected, partially explain sensitivity to right hemisphere
the MAST total, index and sub-test scores were injury. An additional MAST sub-scale (following
highly effective in distinguishing patients with left verbal instructions) involves following instructions
hemisphere stroke from non-patients. that progress in length. This task requires several
The MAST total and index scores were also distin- cognitive abilities including language comprehen-
guished patients with left hemisphere stroke from sion, attention/concentration, left/right discrimina-
those with right hemisphere stroke. The LH group tion and body schema. Cognitive impairments
was more impaired than the RH group on the in these domains (particularly attention) may
MAST total and index scores as well as seven of the have impacted performance on this sub-test for
nine sub-test scores. The two sub-scales for which individuals in the RH group. The writing to dictation
LH group members were not more impaired than sub-scale of the MAST requires motor performance
RH group members were object recognition from a that may be affected by left neglect or other visual-
field of five and verbal fluency. Both of these tasks perceptual impairments. Performance was similar
require the respondent to visually scan stimuli. The for both the RH and NP groups on the sub-
observations indicated that such tasks were difficult scales that were primarily language dependent (i.e.
for patients with left neglect. This may explain the automatic speech, repetition and yes/no accuracy).
failure of these two tasks to differ between left and This initial investigation into the psychometric
right hemisphere patients with stroke and result in properties of the MAST indicated good criterion
less accurate classification of the RH group on DFA. validity. However, this study had many limitations.
As was expected for a language measure, signifi- This study utilized a convenience sample. Subjects
cant overlap was found in MAST performance were individuals who were admitted with either
690 R. Nakase-Thompson et al.

a right or left hemisphere stroke and who also 2. Uzzell BP, Zimmerman RA, Dolinkas CA, et al. Lateralized
received neuropsychological consultation. Individ- psychological impairment associated with CT lesions in head
injured patients. Cortex 1979;15:391–401.
uals with bilateral findings were not included along 3. Tate RL, Lulyham JM, Broe GA, et al. Psychosocial outcome
with individuals not receiving neuropsychological for the survivors of severe blunt head injury: The result form
consultation. Additionally, the three samples were consecutive series of 100 patients. Journal of Neurology,
not equivalent across demographic characteristics Neurosurgery, & Psychiatry 1989;52:1128–34.
(i.e. age and education) which required statistical 4. McClenahan R, Johston M, Densham Y. Misperceptions of
comprehension difficulties of stroke patients by doctors,
control for these factors in examining the psycho-
nurses, and relatives. Journal of Neurology, Neurosurgery, &
metric properties of the MAST. Future research Psychiatry 1990;53:700–01.
should address these limitations and include investi- 5. Kertesz A. Western aphasia battery test manual. The
gation into other aspects of the psychometric proper- Psychological Corporation; 1982.
ties of the MAST. Investigations into the MAST’s 6. Goodglass H, Kaplan E. The assessment of aphasia and
relationships with other measurement tools, utility related disorders. Malvern: Lea & Febiger; 1983.
7. Spreen O, Strauss E. A compendium of neuropsychological
with other patient populations and reliability (i.e. tests. New York: Oxford University Press; 1998.
inter-rater reliability) are indicated. 8. Powell FE, Bailey S, Clark E. A very short version of the
The MAST was developed in response to the need Minnesota aphasia test. British Journal of Social and Clinical
for a brief screening measure that could be adminis- Psychology 1980;19:189–94.
tered at bedside or clinic appointments by a variety 9. Helm-Estabrooks N, Ramsberger G, Morgan AR, et al.
Boston assessment of severe aphasia. Chicago: Riverside;
of health-care providers. Clinicians are frequently
asked to comment on cognitive abilities including 10. Nicholas ML, Helm-Estabrooks N, Ward-Lonergan J, et al.
communication skills and language function. This Evolution of severe aphasia in the first two years post onset.
has implications for implementing medical and Archives of Physical Medicine and Rehabilitation 1993;
rehabilitation interventions, monitoring the course 74:830–37.
11. Benton AL, Hamsher K, Sivan AB. Multilingual aphasia
of recovery of language impairments and providing
examination third edition. Iowa City: AJA Associates; 1994.
feedback to families for education and discharge 12. Enderby P, Crow E. Frenchay aphasia screening test: Validity
planning. The MAST promises to be a useful tool and comparability. Disability and Rehabilitation 1996;18:
in this regard. This brief measure provides a practical 238–40.
method for evaluation of language abilities that 13. Derenzi E, Vignolo L. The token test: A sensitive test to
is amenable to repeated evaluations over time. detect receptive disturbances in aphasic patients. Brain 1962;
14. Kaplan EF, Goodglass H, Weintraub S. The boston naming
Acknowledgements test, 2nd ed. Philadelphia: Lea & Febiger; 1983.
15. Brooks DN, Deelman BG, Van Zomeren AH, et al. Problems
The authors would like to acknowledge the following in measuring cognitive recovery after acute brain injury.
for their assistance in completion of this project: Journal of Clinical Neuropsychology 1984;6:71–85.
Wing Ng, MD, Sid Dickson, PhD, Brook Bible, 16. Still CN, Goldschmidt TJ, Mallin R. A new brief clinical
assessment for aphasia-apraxia-agnosia. Southern Medical
SLP, Robert Adams, PhD, Ashley Doughty, MA,
Journal 1983;76:52–4.
and Candice Harris, BS. 17. Biniek R, Huber W, Wilmes K, Glindemann R, Brand H,
Fiedler M, Annen C. Ein Test zur Erfassung von Sprach- und
References Sprechstorungen in der Akutphase nach Schlaganfallen.
Nervanarzt 1991;62:108–15.
1. Kertesz A, McCabe P. Recovery patterns and prognosis in 18. Available online at: http://www.tbims.org/combi/mast/index.
aphasia. Brain 1977;100:1–18. html

Appendix 1: A sample of MAST individual items

Naming [5 items] (‘What is this called?’)
(1) ______Pen
(2) ______Hand ( point to both sides of your hand )
Automatic Speech [5 items] (‘Finish these sentences for me’)
(1) ______Count to ten (1 ¼ cueing required)
(2) ______I pledge allegiance to the
Repetition [5 items] (‘Repeat these words’)
(1) ______ pot
(2) ______The silver moon hung in the dark sky
Following instructions [5 items]
(1) ______‘Point to your nose’
(2) ______‘Before opening your mouth, touch your ear’
Severe language impairments 691

Yes/no responses [10 items] (‘I’m going to ask some questions; just tell me yes or no’ )
(1) ______Am I touching my eye (clinician touches his/her nose)?
(2) ______Do you put your shoe on before your sock?
Writing/spelling [5 items] (‘Now I would like for you to write some words for me, spell ______’)
(1) ______sit
(2) ______computer
Object recognition in a field of five [5 items] ( ‘I want to show you some things, point to them as I call them out’)
(1) ______Watch
(2) ______Keys
Reading instructions [5 items] (‘Read this aloud and do what it says’)
(1) ______Open your mouth
(2) ______Point to your left ear after you make a fist.
Verbal fluency dictation: [SIDEWALK] write all words that the patient verbalizes and code unintelligible
INSTRUCTION: Show the photo (10 seconds) and say, ‘Look at this picture for a while (pause) Now tell
me everything that you can about this picture; keep talking until I tell you to stop’, immediately start timing
for 10 seconds.