SFBT
SFBT
ORIGINAL
                                                                                                                                                          Fam Med Com Health: first published as 10.15212/FMCH.2017.0118 on 1 December 2017. Downloaded from http://fmch.bmj.com/ on March 18, 2023 by guest. Protected by copyright.
                      Family Medicine and Community Health
                      ORIGINAL RESEARCH
   Furthermore, an important challenge in early identifica-         what the child would have to do, and the approximate time
tion of developmental disability is having tools that respond       required had been explained to them.
to local differences, including cultural perceptions in mean-
ing of childhood ability levels, and that can be used across        Assessment tools
countries [2, 6, 7]. Besides the differences in the factor con-     Seguin Form Board Test
structs, the performances and interpretations in intelligence       In 1856 Seguin developed a simple performance-based intel-
measures differ furthermore between nations and cultural            ligence test using form boards to evaluate eye–hand coor-
settings [8, 9]. While many standardized tools developed            dination, shape concept, visual perception, and cognitive
in Western countries have been validated in their settings,         ability through nonverbal means. It is used to assess the par-
there are many practical limitations in score interpreta-           ticipants’ motor dexterity, visuomotor coordination, spatial
tion and implementation in resource-constrained settings in         organization, and speed and accuracy of performance, and
LMICs [6].                                                          can be used in children as young as 3 years [15, 16]. The
   The conventional intelligence tests predominantly look           form board consists of 10 differently shaped wooden blocks,
at mathematical and verbal intelligences, and are difficult to      and the participants are required to fit the differently shaped
execute and validate in very young children. Moreover, these        blocks into their respective slots on the form board. This
tests require a trained child psychologist, and are arduous in      culture-fair test, which can be easily administered in 10 min,
community settings with respect to their duration, administra-      is used for preliminary assessment of mental age in a normal
tion guidelines, and strict protocols. Consequently, there is a     population.
need for a short, easy-to-use test of ability that can be used as      The task administration involved three consecutive tri-
a quick screening tool in the preschool age of 3–5 years, espe-     als with an instruction to start placement of blocks at the
cially in community settings in LMICs. The current study was        command “Start.” Speed is stressed at the start of the test,
undertaken in this context with the objective to evaluate the       with no further between cues or assistance being provided
concurrent validity for the Seguin Form Board Test (SFBT), a        to the child. The best time from three trials was used to
culture-fair cognitive test, with the Vineland Social Maturity      determine a mental age from the standard chart, which
Scale (VSMS) at 3 years of age and the predictive validity with     was subsequently used in determination of the intelligence
Malin’s Intelligence Scale for Indian Children (MISIC) at 7         quotient (IQ) [17].
years of age.
                                                                    Vineland Social Maturity Scale
Materials and methods                                               The VSMS, originally developed by E.A. Doll in 1935, was
The present study is part of a wider study of a birth cohort of     adapted to the Indian scenario by Malin [18]. It is a semistruc-
452 children from a semiurban slum area in Vellore, South           tured assessment based on an interview of the caregiver and
India [10], in whom diarrheal episodes were being followed up       evaluates the social ability of the child.
for 3 years for studies on rotaviral [11–13] and cryptosporidial       In this questionnaire method, adaptive functioning is
[14] diarrhea. The study is designed as a panel study using         measured in the context of self-help skills, self-direction,
a modified longitudinal design that performed cross-sectional       socialization, and communication. The measure has eight
analyses for intelligence and social maturity at 3 and 7 years of   social domains with 89 items, and can be used from birth
age in part of a birth cohort of children to evaluate long-term     to 15 years. The VSMS has good concurrent validity of
effects on cognition of early childhood diarrhea. A subsample       at least 0.8 with intelligence tests in children with mental
of 116 children enrolled in the birth cohort underwent intel-       retardation [19, 20]. The maturity age associated with the
ligence and social maturity assessments at 3 years of age, and      level of functioning was calculated along with a social matu-
291 children were assessed at 7 years of age. Written consent       rity age that is then converted to an index called the ‘social
was obtained from the parent/s about the assessment tools,          quotient’ (SQ).
                                                                                                                                                    Fam Med Com Health: first published as 10.15212/FMCH.2017.0118 on 1 December 2017. Downloaded from http://fmch.bmj.com/ on March 18, 2023 by guest. Protected by copyright.
                     Koshy et al.
           Malin’s Intelligence Scale for Indian Children                      data for socioeconomic status and sex. Concurrent validity was
           MISIC, the Indian adaptation of the Wechsler Intelligence           analyzed by evaluation of the relationship between IQ and SQ
           Scale for Children [21], measures verbal and performance            at 3 years of age with use of Pearson’s correlation coefficient
           abilities, and can be administered by a trained psychologist to     (r). Predictive validity was calculated with use of Pearson’s
           children to assess intelligence from 6 years onward [22–24].        correlation coefficient between IQ at 3 years of age and IQ at 7
           The intelligence scale measured as the full IQ is obtained from     years of age. Partial correlation coefficients were calculated to
           six verbal subscales and five performance subscales. The verbal     adjust the data for the effect of sex, socioeconomic status, and
           scale measures verbal information and language development          birth weight on the relationship between scores. A P value of
           and comprehension, using the following subtests: information,       less than 0.05 was considered statistically significant. All anal-
           similarities, arithmetic, vocabulary, comprehension, and digit      ysis was done with STATA 10.0 (StataCorp, College Station,
           span. The performance scale has the following subtests: pic-        TX, USA).
           ture completion, coding, picture arrangement, block design,
           and object assembly. The raw scores obtained from the ver-          Results
           bal subscales are converted into standardized scores to derive      For the 3-year assessment, 116 children were evaluated at an
           the verbal IQ. Similarly, the performance subscales yield the       average age of 3.46 years (standard deviation 0.37 years). The
           performance IQ, and the cumulative score of the verbal and          7-year assessment included 291 children at an average age
           performance subscales gives the full-scale IQ. The test has a       of 7.23 years (standard deviation 0.38). Ninety-five children
           reliability coefficient of about 0.9, and the concurrent and con-   underwent both the 3-year assessment and the 7-year assess-
           gruent validity scores are both around 0.6 [16, 17].                ment (Fig. 1). Table 1 shows the sociodemographic profile of
                                                                               the 95 children. The children studied are representative of the
           Procedure                                                           total study population.
           Institutional ethics committee clearance was obtained for
           this study, and the institutional review board approved both
           assessment protocols. The 3-year assessment was performed
                                                                                          3 years                          7 years
           between June and December 2005, and the 7-year assessment
           was performed between December 2009 and April 2010.
              For the purpose of this study, all measures were translated
           to the local language, Tamil, and back-translated, and a pilot
           study was conducted for appropriateness of measures and                    116 recruited                      291 recruited
           items, before the commencement of individual assessments.
           At 3 years a medical research officer trained in testing evalu-
           ated the children with the SFBT and the VSMS. At 7 years a
           child psychologist assessed the children using MISIC and the                21 did not                      291 completed
           VSMS. Both assessments were conducted in a distraction-free               complete SFBT                        MISIC
           environment in a separate quiet room in the field clinic. The
           respective intelligence and social maturity tests were com-
           pleted on the same day. The assessment of socioeconomic sta-
                                                                                                                              95 included for
           tus was performed with the modified Kuppuswami scale [25].                        95 included for
                                                                                                                               analysis who
                                                                                                 analysis
                                                                                                                               took 3 years
                                                                                                                                assesment
           Statistical analysis
           All study variables were summarized with use of descriptive         Fig. 1. Flowchart depicting numbers of participants in both
           statistical methods. Analysis was done after adjustment of the      assessments
Table 1. Comparison of sociodemographic characteristics between the study population and the birth cohort nonparticipants in the study
χ test.
a 2
b
    Two-tailed t-test. The mean is given, with the standard deviation in parentheses.
Mann-Whitney U test. The median is given, with the interquartile range in parentheses.
c
d
    Data not available for 11 children.
       For the SFBT at 3 years of age, 21 children could not complete            score (SQ) of 121.1 (t=-3.05, P=0.002). The 7-year analysis
the test within the stipulated time. The internal consistency of the             showed a low normal mean intelligence (IQ) of 85.5, significantly
three trials for the SFBT was high at 0.906. The 3-year analysis                 lower than the corresponding average social maturity quotient
showed an average mean intelligence (IQ) of 110.7 on the SFBT,                   (SQ) of 101 (t=-8.97, P<0.001). Table 2 summarizes the IQs and
significantly lower than the corresponding mean social maturity                  SQs for children assessed at both 3 and 7 years of age.
Table 2. Summary of intelligence and social quotients in the children assessed at both 3 and 7 years (n=95)
Intelligence                   SFBT (IQ)                110. 7             18.0             MISIC, full scale (IQ)             85.5           11.6
                                                                                            MISIC (VQ)                         84.0           11.1
                                                                                            MISIC (PQ)                         82.7           15.2
Social maturity                VSMS (SQ)                 121.3             28.7             VSMS (SQ)                        101.0            12.2
IQ, intelligence quotient; MISC, Malin’s Intelligence Scale for Indian Children; PQ, Performance intelligence quotient; SD, standard deviation;
SFBT, Seguin Form Board Test; SQ, social quotient; VQ, Verbal intelligence quotient; VSMS, Vineland Social Maturity Scale.
                                                                                                                                                    Fam Med Com Health: first published as 10.15212/FMCH.2017.0118 on 1 December 2017. Downloaded from http://fmch.bmj.com/ on March 18, 2023 by guest. Protected by copyright.
                     Koshy et al.
           Concurrent validity for the SFBT                                       Other comparative studies have also reported a simi-
           IQ had a moderate positive correlation with SQ at 3 years           lar finding of social score being better than the intelligence/
           of age (r = 0.39, P= 0.04). After we had controlled for the         developmental score. Bhave et al. [27, 28] showed that signifi-
           effects of sex, socioeconomic status, and birth weight, the         cantly higher mean SQs than mean developmental quotients
           adjusted correlation between IQ and SQ at 3 years of age            were obtained on the same set of children. Song and Jones
           was 0.38 (P= 0.04). For comparative purposes, the adjusted          [29] reported an overestimation of social age by 1–2 years with
           correlation between IQ and SQ at 7 years of age was 0.40            the VSMS in normal children. However, studies by Raggio
           (P<0.001).                                                          et al. [30] have shown comparable SQs and developmental
                                                                               quotients.
           Predictive validity for the SFBT                                       This study shows that there is a positive correlation
           There was a moderate positive correlation between the IQs           between IQ and SQ at 3 and 7 years of age. The strength of the
           measured by the SFBT and MISIC. Comparing the analysis              relationship was comparable at both 3 and 7 years of age. This
           at three and 7 years of age, the adjusted Pearson’s correlation     is an important finding showing not just concurrent corrobo-
           coefficient for IQ was 0.66 (P<0.001), after we had controlled      ration but also stable properties of the intelligence construct
           for the effects of sex, socioeconomic status and birth weight.      measured by the SFBT to later years, both valuable properties
           IQ at 3 years correlated with both Verbal IQ (r=0.57, P=0.001)      essential in intelligence assessments in community settings.
           and Performance IQ (r=0.54, P=0.003).                                  Indian studies have shown a good correlation of social abil-
                                                                               ity with intelligence tests in children with intellectual disabil-
           Discussion                                                          ity or mental retardation [19]. Although social adaptability is
           The results of the study suggest that the SFBT has moder-           different from intelligence, the VSMS can be used in resource-
           ate concurrent validity with the VSMS and moderate predic-          poor settings to pick up deviations, enabling the health worker
           tive validity with MISIC at a later age. This tool can evolve       to refer the child for further evaluations. However, the VSMS,
           as an effective community-based intelligence-testing tool in        as with other parent-report questionnaires, is a subjective
           resource-limited settings. The brevity of the test, its portabil-   interpretation of ability. The Vineland Adaptive Behavioral
           ity, the ability to arouse attention or sustain interest, and the   Scales (VABS) [31], an extensive revision of the VSMS, may
           ease of administration are some of reasons that the SFBT lends      be ideal to calculate social maturity, but is yet to be adapted to
           itself to be used as a screening tool before further referral for   the Indian scenario. Further studies are required to standard-
           early learning interventions [26]. To the best of our knowledge,    ize the VABS for resource-limited settings and its correlations
           this is the first study attempting to correlate an early assess-    with the VSMS and other intelligence/developmental tools.
           ment tool such as the SFBT with intelligence testing at a later        Some studies have evaluated the SBFT norm in Indian
           age in resource-poor settings. The current study enhances the       children. Basavarajappa et al. [32] found that the SFBT con-
           already existing literature by presenting good concurrent, and      tinues to remain a valid and reliable speed test of intelligence
           predictive validity for this measure in the community setting.      in younger children. However, they advocate separate SFBT
              The children included in the present study are representa-       norms to account for differences in age, sex, socioeconomic
           tive of the cohort population and consisted of an almost equal      level, and residential setting (rural/urban). Thangavel [33]
           number of boys and girls. The mean SQs at both 3 and 7 years        found that sex tends to influence the performance in children.
           of age are higher than the corresponding mean IQs. The dif-         In 1968 Ramachandran et al. [34] found that Indian children
           ference was statistically significant at both 3 years (P= 0.002)    had slower performance speed than their Western counter-
           and 7 years (P<0.001) of age. The children examined in our          parts. Contrarily, in 1971 Bharatraj [35] reported that a sample
           study scored well on the self-help items of the VSMS, giv-          of Mysore children studied were on average faster than chil-
           ing them a better score than the corresponding intelligence         dren in other reports. Goel and Bhargava [36] replicated this
           scores.                                                             result in their study with a sample of Delhi school children
aged between 3 and 15 years. Verma et al. [37] reported com-        Significance statement
paratively better performance in speed of same-aged children        The Seguin-Form Board Test (SFBT) is an intelligence testing
from upper-class schools. A study by Venkatesan [38] indi-          tool that is brief, portable and easy to administer, highlighting
cated that three trials might not be sufficient to determine        its utility as a community screening tool in resource-limited
the mental age equivalence of the child. According to that          setting. The present study demonstrates acceptable concurrent
study, the optimum performance is observed at the sixth trial.      validity for SFBT as early as three years of age as well as stable
Venkatesan also computed a ‘decrement score and quotient’           properties of the construct with acceptable predictive validity
that showed an inverse relationship between increasing cal-         with intelligence assessment at seven years of age, both valuable
endar age of the child and the time taken to complete the test.     properties of intelligence assessments in community settings.
   Although the SFBT is easy to administer and takes a maxi-
mum 10 min, it is reported to be limited in measuring only the      References
performance intelligence such as visuomotor coordination and        1.   UNICEF, WHO. Early childhood development and Disability
does not assess verbal skills [16, 39]. However, the intelligence        [Internet]. 2012 [cited 2017 Feb 20]. Available from: www.
                                                                         scribd.com/document/297962893/Early-childhood-develop-
score from the SFBT in our study had moderate positive and
                                                                         ment-and-Disability-Unicef-Who-2012.
comparable correlations with verbal and performance scores at
                                                                    2.   Sabanathan S, Wills B, Gladstone M. Child development
7 years of age. Nevertheless, because of the very young age at
                                                                         assessment tools in low-income and middle-income countries:
the initial assessment, a significant number (18%) of children
                                                                         how can we use them more appropriately? Arch Dis Child
did not cooperate or complete the tests during the stipulated            2015;100(5):482–8.
time and could not be included in the final analysis. Despite       3.   Grantham-McGregor S, Cheung YB, Cueto S, Glewwe P, R
                                                                                                                             ichter
these limitations, the present study establishes the prospect of         L, Strupp B. Developmental potential in the first 5 years for chil-
using the SFBT as an intelligence tool in children as young as           dren in developing countries. Lancet 2007;369(9555):60–70.
3 years in resource-limited settings. The utility of this test in   4.   Gardner H. Frames of mind: the theory of multiple intelligences.
the community is highly promising, where a trained person                Review by: Blythe McVicker Clinchy. J Educ 1984;166(2):
can screen the ability of the child before further referral.             199–205.
   This study analyzed children at a very young age of 3            5.   Goldberg Hermo X, Lemos Giráldez S, Fañanás Saura L. A sys-
years, an age when most cognitive assessment tests cannot be             tematic review of the complex organization of human cognitive
                                                                         domains and their heritability. Una revisión sistemática de la
performed. Despite the floor effects of the test, most of the
                                                                         organización compleja de los dominios cognitivos humanos y
study children completed the SFBT. The utility of the SFBT as
                                                                         su heredabilidad [Internet]. 2014 [cited 2017 Feb 21]; Available
a school-readiness tool in very young children starting kinder-
                                                                         from: http://digibuo.uniovi.es/dspace/handle/10651/25306.
garten or school in LMICs needs to be explored further, as most
                                                                    6.   Fischer VJ, Morris J, Martines J. Developmental screening tools:
schools in these settings do not use any scientific measures
                                                                         feasibility of use at primary healthcare level in low- and middle-
to assess the abilities of such children. Prompt identification          income settings. J Health Popul Nutr 2014;32(2):314–26.
of additional needs at school entry can help schools, teach-        7.   Peña ED. Lost in translation: methodological considerations in
ers, families, and children themselves to optimize resources to          cross-cultural research. Child Dev 2007;78(4):1255–64.
accomplish the best learning potential for each child.              8.   Rosenqvist J, Lahti-Nuuttila P, Urgesi C, Holdnack J, Kemp
                                                                         SL, Laasonen M. Neurocognitive functions in 3- to 15-year-old
                                                                         children: an international comparison. J Int Neuropsychol Soc
Conflicts of interest
                                                                         2017;23(4):367–80.
The authors have no conflicts of interest.
                                                                    9.   Thompson B, Kirby S, Smith K. Culture shapes the evolution of
                                                                         cognition. Proc Natl Acad Sci USA 2016;113(16):4530–5.
Financial support                                                   10. Gladstone BP, Muliyil JP, Jaffar S, Wheeler JG, Fevre AL,
The cohort study was supported by Wellcome Trust grant no.               Iturriza-Gomara M, et al. Infant morbidity in an Indian slum birth
063144.                                                                  cohort. Arch Dis Child 2008;93(6):479–84.
                                                                                                                                                                Fam Med Com Health: first published as 10.15212/FMCH.2017.0118 on 1 December 2017. Downloaded from http://fmch.bmj.com/ on March 18, 2023 by guest. Protected by copyright.
                      Koshy et al.
           11. Banerjee I, Gladstone BP, Le Fevre AM, Ramani S, Iturriza-             25. Kuppuswamy B. Manual of socioeconomic status (urban). In:
               Gomara M, Gray JJ, et al. Neonatal infection with G10P[11]                 manual of socioeconomic status (urban). 1st ed. Delhi: Mana-
               rotavirus did not confer protection against subsequent rotavirus           sayan; 1981. pp. 66–72.
               infection in a community cohort in Vellore, South India. J Infect      26. Newell DC. The uses of the form board in the mental measure-
               Dis 2007;195(5):625–32.                                                    ment of children. Psychol Bull 1931;28(4):309–18.
           12. Ramani S, Banerjee I, Gladstone BP, Sarkar R, Selvapandian D,          27. Bhave A, Bhargava R, Kumar R. Correlation between develop-
               Fevre AML, et al. Geographic information systems and genotyp-              mental quotients (DASII) and social quotient (Malin’s VSMS) in
               ing in identification of rotavirus G12 infections in residents of an       Indian children aged 6 months to 2 years. J Paediatr Child Health
               urban slum with subsequent detection in hospitalized children:             2011;47(3):87–91.
               emergence of G12 genotype in South India. J Clin Microbiol             28. Bhave A, Bhargava R, Kumar R. Development and validation of
               2007;45(2):432–7.                                                          a new Lucknow development screen for Indian children aged 6
           13. Banerjee I, Ramani S, Primrose B, Moses P, Iturriza-Gomara M,              Months to 2 Years. J Child Neurol 2010;25(1):57–60.
               Gray JJ, et al. Comparative study of the epidemiology of rotavi-       29. Song AY, Jones SE. Vineland Social Maturity Scale norm exam-
               rus in children from a community-based birth cohort and a hospi-           ined – the Wisconsin experience with 0- to 3-year-old children.
               tal in South India. J Clin Microbiol 2006;44(7):2468–74.                   Am J Ment Defic 1982;86(4):428–31.
           14. Ajjampur SSR, Gladstone BP, Selvapandian D, Muliyil JP, Ward           30. Raggio DJ, Massingale TW. Comparison of the Vineland Social
               H, Kang G. Molecular and spatial epidemiology of cryptosporid-             Maturity Scale, the Vineland Adaptive Behavior Scales – survey
               iosis in children in a semiurban community in South India. J Clin          form, and the Bayley Scales of Infant Development with infants
               Microbiol. 2007;45(3):915–20.                                              evaluated for developmental delay. Percept Mot Skills 1993;
           15. Cattell RB. A guide to mental testing. 3rd ed. London: University          77(3 Pt 1):931–7.
               of London Press; 1953.                                                 31. Becker-Weidman A. Effects of early maltreatment on develop-
           16. Venkatesan S. Celebrating a century on form boards with special            ment: a descriptive study using the Vineland Adaptive Behavior
               reference to Seguin Form Board as measure of intelligence in               Scales-II. Child Welf 2009;88(2):137–61.
               children. Glob J Interdiscip Soc Sci 2014;3(6):43–51.                  32. Basavarajappa S, Venkatesan D, Vidya M. Normative data on
           17. Arya S. Manual for intelligence testing. Secunderabad: National            Seguin Form Board test. Indian J Clin Psychol 2009;35(2):93–7.
               Institute for the Mentally Handicapped; 1981.                          33. Thangavel N. A study of sex differences on Seguin Form Board
           18. Malin AJ. Indian adaptation of the Vineland Social Maturity                Test. Child Psychiatr Q 1986;19(4):158–60.
               Scale. Lucknow: Indian Psychological Corporation; 1971.                34. Ramachandran KV, Deshpande VA, Apte, SV, Shukla NN,
           19. Kumar I, Singh A, Akhtar S. Social development of children with            Shah MR. A survey of school children in Mumbai city with spe-
               mental retardation. Ind Psychiatry J 2009;18(1):56.                        cial reference to their physical efficiency, mental and nutritional
           20. Louis PT, Emerson IA. Evaluating the cognition, behavior, and              status. New Delhi, India: Asia Publishing House; 1986.
               social profile of an adolescent with learning disabilities and         35. Bharatraj, J. AIISH norms on Seguin Form Board with Indian
               assessing the effectiveness of an individualized educational pro-          children. J All India Inst Speech Hear 1971;2:117–27.
               gram. Iran J Psychiatry Behav Sci 2014;8(2):22–37.                     36. Goel SK, Bhargava M. Handbook for Seguin Form Board. Agra:
           21. Wechsler D. The wechsler intelligence scale for children. New              National Psychological Corporation; 1990.
               York: Psychological Corporation; 1949.                                 37. Verma SK, Pershad D, Randhawa A. A preliminary report on a
           22. Malin AJ. Malin’s intelligence scale for children – manual. Luc-           performance test of intelligence on 4-8 year old children. Indian
               know: Indian Psychological Corporation; 1969.                              J Clin Psychol 1979;6(2):125–30.
           23. Mahakud GC. Dyslexia: an introduction to learning disorder.            38. Venkatesan S. Psychological assessment of rural children with
               New Delhi: Tata McGraw-Hill Education; 2013.                               mental handicaps in India: some problems and issues. Creat Psy-
           24. Rakhee KJ, Oommen A, Shobini LR. Neuropschylogical pro-                    chol 1991;7(1 & 2):1–9.
               file of learning disability in the Indian population-preliminary       39. Mittler P. The use of form boards in developmental assessment.
               results. Indian J Clin Psychol 2007;34(1):70–4.                            Dev Med Child Neurol 1964;6:510–6.