The relationship between stuttering and academic achievement in children

The relationship between stuttering and academic achievement in children

~n~~u~tion The WS~HSC~of a p&ble relationship between stuttering and academic sehiwemcslnt in duldran has been of interest to investigators over the ...

928KB Sizes 1 Downloads 113 Views

~n~~u~tion

The WS~HSC~of a p&ble relationship between stuttering and academic sehiwemcslnt in duldran has been of interest to investigators over the past halfstudy (~CWadi IYl2) ftlund stutterers to bc than their cfsssmates, an apparent indication that the child who starttess f~ one hal@adr! retarded in relation to nanstuttering i;hildren of the game ftgp A study by Raot (¶926] pr0duce4 lho same finding. These s;tudiee,howlwor, said littie &bout the way stutterers f‘unction within their classw36mtt. Since; that which we measure as *‘academicachievement” is more clasely tied to ~yatmafk progress through the grades than to age (Lindquist and Hieronymus 1964), it is Newell to exsrnine the comparative academic achievement of st~tte~e~~and n~~~t~~ttffr~~~ who havrtthe same grade placement, rather than age dlffmnces between thg two groups in any specific grade. Standardized t-sts of achievement provide a tosl f’r making these comparisons. MclI?~waI(1928) employed tha Stanford Achievement Test fox redding, a&hmetic and spelling in an examination of the acsdemic achievement of a ~QUP of 45 stuttering children (of unspecified age and grade placement) and a gsaupof nonstuttcrers equivalent “in chronological age, mental age, intelligence

century. $ha ear&+t

fepertcd

about ~~~~~~~ yeinr alder

D.E.WILLIAMS, B.M.MELROSE and C.L.WOODS

88

(as determined by the Stanford Revision of the Binet Inteliigence Test), Sex, and language and racial background”. The tests showed a slight superiority for the nonstutterers in every case, although the results were statistically not significant. Du Pant (1946) obtained achievement test scores on 126 stuttering children in Fades ace throu@~ twelve. The tests used varied for the various age lids. and included the Iowa Tests of Basic Skills, the Metropolitan Primary ‘tests, and the IQwa Silent Reading Test. No control group was used, but the stutterers’ scores were significantly lower than the published norms in langu grades three through five, and in basic arithmetic skills in grades eight. However, this finding must be interpreted with caution, sin able difference in academic achievement test scores exists between boys and girls, especially in language skills (Lindquist and Wieronymus 1964), and this finding is based on the comparison of a predominantly male stuttering grcjup with normative data combining scores from both sexes. It is important to determine whether a relationship exists between stuttering and academic achievement in children, since academic achievement is generally acknowledged to influence many areas of adult life, including vocational choice. This paper presents the results of two studies designed to investigate this relationship. The purpose of Investigation I was to examine differences in academic achievement in grade six between a group of stutterers and a group of nonstutterers. The purpose of Investigation II was to compare changes in academic achievement level which occurred from grade four to grade eight in a group of stutterers and a group of nonstnttterers. Both investigations utilized the Iowa Tests of Basic Skills, which are recognized to be among the better and more widely used tests of achievement in the elementary grades (Buros 1959). The tests yield a score or scores for each of five major areas: Vocabulary, Reading. Language (four subtests and a total), Work-Study Skills (three subtests and a total), and Arithmetic (two subtests and a total). In addition, a single cornposite score is provided for each child. The score yielded by each of the measures is expressed in terms of grade equivalent units *, and tables are provided for transforming grade equivalent scores into percentile ranks, A large number of school systems in Iowa, where the investigations were carried out, routinely administer these tests in grade four, six and eight, and all of the test results used were obtained from these routine testings in either public or parochial schools. The school systems from which test results were obtained represented both rural and urban areas, and ranged from small to large in enrollment. The great majority of the stuttering children whose sc)res were used had been * The grade equivalent score indicates the grade level at which the typical pupil makes B given raw score. The first digit represents the grade and the second digit the month within the grade.

ACADEMIC

.\CHIEVEMENTINSTldTTERINGC)iILDREN

89

identified as such through routine referral or screening procedures in the schools, and their names submitted by local speech clinicians. Nonstuttering control children in lnve~ti~at~on I were selected at random from the same sixth-grade ~~a~r~~s as the stuttering children. Nonstuttering control children in Investi~~t~o~ II, with very few exceptions, were selected from the same fourth-grade oom as stutterin children and were in the same class and school at grade Three nonstutt ing children were used to each stuttering child to insure a mote f~~re~~tat~ve control group population. The method of selecting controls was felt to he adequntp to make very unlikely any substantial differences in either s~~joe~or~orni~ or ed~l~ationa~ background in the two groups.

tion I w&s des&jned to answer the following questifJn: I>0 stuttering children in tke Gxth grade differ from their nonstuttering classmates in either overall a~adern~~ ~~i~~ev~rn~ntor in specific areas of study’? The WSS scores of 100 sixth-grade stutterers and 300 sixth-grade nonstuttering controls were used in this investigatian, and the ratio of male tc female subjects was 6.7 : 1. y polygons were plotted from the score distributions of the two , l-6). The following observations can be made from inspection of these polygons: (1) there is considerable overlap of the score distributions of stutterers and nonstutterers on all measures; (2) the range of scores for both groups is essentially similar; (3) the disrilbutions of stutterers’ scores on some measures appear to be skewed to the righht.

---

stutteren

-

Nonrtuttrrers

0 GRAS

EQUIVALENT

SCMES

Fig. 1.Vocabuiary skills.

et

- - -Stuttrrora -Nonatuttrron

,4

(L

24 c

f I

NW

\ \

EOUIVAL EN1 SCaCIES Fig. 2. Reading skills.

6RAOE

- - - stutteraro Nonrtuttsm

GRAM

EWVALENT SCORES

Fig. 3. Language skills. -- -Stuttawr -NonotuWuem

24 .

~m2 Fig. 4.

EQUIVALENT smtx3

Work-study skills.

ACADEMIC ACHIEVEMENT IN STUTTERING CHILDREN

30-

91

- - - Stutteras

wnstuttaen

-

2?_ 24 a

VI

\\ \’

--&T--e* GRADE

EQUIVALENT

Fig. 5. ArithnetLi:

skills. - - -sturterem Nonstvtterwn

30 27

;\ I ’\

_

: I / I 1 *1’ I

24 -

E $

2’18.

3 ~

15-

: /

SCORES

i

\ \ * \ \ \ \ \

0

[‘f

f

.

‘!y~:,

t

\

\ 5

GRADE

EOUIVALENT

SCORts

Fig. 6. Composite.

Median score values for the two groups are presented in table 1. The median was chosen as the preferred measure of central tendency because it is felt to be more representative of a “typical” score for each group than is the mean. It can be seen from table 1 that for all test measures the median score of the nonstutterers is higher than that of the stutterers. For only two measures, spelling and language usage, is this difference not statistically significant. The findings of this investigation demonstrate consistent academic retarda-

D.E.BXLIAMS, B.M.MELROSE and C.L.WOODS

92

Table 1 Median grade equivalent scores of 100 sixth-grade stutterers and 300 sixth-grade nonstuttering :ontrol subjects on the various items of the IOwa Tests of Basic Skills. Differences are reported in months. Median --.

Item

Nonstu tterers

- -Stutterers

Difference

_-

-.-

--

Vocabulary (V)

66.5

59.0

7.5 *

Reading (R)

65.0

61.2

3.8 *

Language (L) Spelling Capitalization Punctuation Usage

62.3 62.0 62.6 62.5

58.5 57.9 58.8 56.5

3.8 4.1 * 38* 6.0

Total

63.7

57.5

6.2 *

Work-study (W) Maps Graphs References

66.8 64.0 64.5

61.9

4.9 *

60.0 59.2

4.0 * 5.3 *

Total

64.8

61.5

3.3 *

Arithmetic (A) Concepts Problems

66.4 64.5

60.5 60.5

4.0 *

Total

64.8

61.5

3.3 *

Composite (C)

65.6

59.3

6.3 *

5.9 *

* Significant at the .OS level (Mann-Whitney U-test for median differences).

tion for sixth-grade stutterers, as a group, as compared with sixth-grade nonstutterers. However, the fact that the range of scores is approximately the same for both groups indicates that this finding is not necessarily true for any individual sixth-grade stutterer, whose score may place him anywhere from the top to the bottom of the total range of scores. it can be said of any sixth-grade stutterer only that he is more likely to be academically retarded than is any nonstutterer.

Investigation II

Investigation 11was designed to answer the following question: does progress in academic achievement for stutterers differ from that for nonstutterers over the period from grade four to grade eight? The fourth- and eighth-grade ITBS scores of SOmale stuttere, Lbmd 150 male

ACADEMIC ACHIEVEMENT IN STUTTERING CHILDREN

93

Table 2 Median grade equivalent scores of 50 fourth-grade stutterers and 150 fourth-grade nonstuttering control subjects on the various items of the Iowa Tests of Basic Skills. Differences are reported in months (positive differences favor ncnstutterers). Median Item

-11 Nonstutterers

Stl;tterers

Difference

Vocabulary (V)

47.1

43.5

3.6

Reading (R)

46.8

42.4

4.4 *

Language (L) Spelling Cspitaliza t ion Puirctuation Usage Total

46.5 47.2 48.3 43.1 47.5

43.5 45.2 43.8 38.0 41.5

3.0 2.0 * 4.5 5.1 * 6.0 *

Work-study (W) Maps Graphs References Total

48.9

45.5 44.4 46.7

45.2 43.5 45.0 43.3

3.7 * 2.0 -0.6 3.4

Arithmetic (A) Concepts Problems Total

48.0 45.0 45.8

45.2 45.5 44.8

2.8 -0.5 1.0

Composite (C)

47.0

42.0

5.0 *

* Significant at the .OSIcvcl (Mann-Whitney U-test for median differences).

nonstuttering controls were used in this investigation. Twelve of the stutterers whose scores were used were children whose sixth-grade scores were used in Investigation 1. The remaining 38 sets of scores were obtained from children who had not served as subjects in Investigation I. In order to make direct comparisons of data from the fourth and eighth grades, the grade equivalent scores for all subjects were converted to percentile ranks. It could then be determined whether each subject’s rank went up or down from grade fi>ur to grade eight. For the group of stutterers, there was no measure on the test for which a significant rise or drop in percentile rank was found. For the group of nonstutterers, there was no measure for which a significant rise was found. However, there were three measures, Spelling, Punctuation and Language Total, for whi& a significant * drop was found. The first two of these measures are sub* 0.05 level of significance, Sign test (Siegel 1956).

D.E.WILLIAMS,B.M.MELROSEand C.L.WOOD,C

94

Table 3 Median grade equivalent scores of 50 eighth-grade stutterers and 150 eighth-grade nonstutteringcc~ltrolsubjects on the various items of the Iowa Tests of Basic Skills. Differences are reported in months (positive differences favor nonstutterers). Median Nonstutterers

Stutterers

M-p DifferenlZe

Vocabulary (V)

86.2

81.0

5.2

Reading (R)

8d.7

:2.8

5.9

Language (L) SpeBing Capitaltiation Punctuation Usage Total

81.3 87.2 86.3 83.5 82.5

81.5 82.5 79.5 78.0 83.8

-0.2 4.7 6.8 5.5 -1.3

Work-study (W) Maps Graphs References Total ‘.

91.4 86.9 85.1 88.3

85.0 88.0 82.0 84.5

6.4 -1.1 3.1 3.8

Aiihmetic (A) Concepts Problems Total

90.0 87.5 87.9

81.0 87.7 83.5

9.0 * -0.2 4.4

Composite (C)

85.8

81.5

4.3

Item

-

* Significant at the .OSlevel (Mann-Whitney U-test for median differences).

of the third. It shou!d be noted that the findings presented are for the total groups, and that chang>s in rank occurred on various measures for individuals within both groups. When the stutterers and nonstutterers wpte compared for differences between the two groups, a significant * difference was found for only the Language Total measure, where 17% more of the nonstutterers than of the stutterers demonstrated a drop in percentile rark The single finding of interest from thi:; investigation of percentile rank changes over a four-year period is that the stutterers, as a group, did not demonstrate the drop in achievement level in language skills which was demonstrated by the nonstutterers. Since Lindquist and Hieronymus (1964) presented evidence of an increasing superiodty of girls over boys in language skills through the elementary years, the drcp in this area demonstrated by the tests

* 0.05level of significance, X2 test (Siegel 1956).

ACADER4IC ACHIEVEMENT IN STUTTERING

CHILDREX

95

gfcrUpof male nonstutterers is not unexpected. It would have been expected, however, that the group of male slutterers would demonstrate the same drop. Median grade equivalent scores for stutterers and nonstutterers are presented in table 2 for grade four and table 3 for grade eight. At grade four, median scores for stutterers me lower than median scores for nonstutterers for all measuresexcept two subtests. At grade eight, median scores for stutterers are lower than those for nonstutterers except for the Language Total measure and the three subtests, Spelling, Graphs and Arithmetic Problems. The fact that a number of tfrese differences are found to be significant at grade four in spite of the relatively small number of subjects used, pravldes strong confirmation for the findings of Investigation I. it is Fossible that the greater variability in scores at grade eight accounted for the fewer significant differences found, since for ten of the fifteen measures the median differences are of greater magnitude than at grade four.

Considerationof the findings of these two investigations into the academic achievementof youngstutterers yields the following conclusions: (1) As a group, stutterers appar to demonstrate a consistent academic retardation, both in overd academic achievement and in special areas of study. Jt should be noted,

however, that although a stutterer is more likely to be somewhat academically retarded than one of his nonstuttering classmates, he is not necessarily so, since the range of achievement test scores for stutterers is almost identical to that for nonstutterers. (2) Far all areas except language skills, there appears to be no difference in the amount of academic achievement made by stutterers and nonstutterers, as groups, between grade four and grade eight. For the Language Total measure not only is there a significant difference in the number of stutterers as compared with nonstutterers who drop in percentile rank, hut the difference of six academic months which existed at grade four between the median stutterer and nonstutterer had dropped to 1.3 months at grade eight. In vjew of these Endings, it appears reasonable to conclude that in the area of language skills, stutterers tend to “catch up” to their nonstuttering peers.

There are aweral factors which lrlay cause or contribute to the academic retardation of your64stutterers observed in these studies. Among the factors to bo considered am the following: (1) Do stutterers, as a group, demonstrate a iower level of native intelligence than nonstutterers? (2) Was the academic development of stutterers, as a group, been influenced sufficiently by environmental factors to account for the demonstrated academic retardation? (3) Is the language development of very young stutterers less adequate than that of non-

96

D.E.WiLLIAMS, B.M.‘MELROSE and C.L.WOODS

stutterers, causing them to experience difficulty in the early years of school? The first and most obvious question to consider is whether young stutterers as a group are handicapped by a lower-than-normal degree of native intelligence. A very limited amount of information is available on the intelligence of children who stutter, all of it over 19 years old. Johnson (1959) reported that no significant difference was found between the distribution of IQ’s of 50 stuttering children and the distribution of IQ’s of the general population as reported by Wechsler. West (1931) found the median IQ for 4,059 stutterers to be 96.5, and DUpant (1946) using the Otis Intelligence Tests with a group of 126 stutterers, found the mean and median IQ, respectively, to be 94.9 and 98.5. The fact that environment has been found to influence IQ test results has brought about a change in the interpretation of these results in recent years. The IQ test is viewed today as a measure of current level of functioning, which is the result of native intelligence plus the effects of environmental influences (Dyer 1964; Eels 1964). When the results of the studies reported above are interpreted in his fashion, it can be stated that any difference in IQ found between stutterers and nonstutterers reflects a difference in level of functioning, but cannot be viewed as evidence of a difference in native intelligence. The function of environmental influences on academic achievement in stuttering children can be considered from several viewpoints. One of the most frequent effects of stuttering is that the amount of the stutterer”s participation in classroom activities is decreased, both because of his own reluctance to speak and because of the teacher’s reluctance to cause him discomfort or embarrassment by calling upon him to speak. One result of this lack of active participation in classroom affairs may be decreased learning. The child who sits passively throughout a class discussion is most probably not learning as much as the child who actively follows the course of the discussion in order to make some contribution. A second manner in which environmental factors may contribute to decreased academic achievement is through the development of test anxiety. Sarason et al. (1964) found a strong negative relationship between test anxiety as measured by the Test Anxiety Scale for Children, and scores on achievement

tests. They hypothesized that the pattern of defenses developed by the child against the anxiety have an interfering effect on performance. They also hypothesized that the cognitive consequences of anxiety can affect performance even though the anxiety is no longer present. In other words, anxiety is an etiological factor restricting intellectual and academic development, Samson et al. (1960) hypothesized that test anxiety develops in test-like experiences both in and out of the classroom, and that the family situation is the most significant out-ofschool situation in which the child’s performance is evaluated by adults whose judgements are important to him. The test-anxious child carries over to the teacher and the SCHOOL his reactions to the evaluations of his behavior made in

ACADEMK ACHIEVEMENT IN STUTTERlNG CHILDREN

97

the home- In this *f@e, the term “evaluational anxiety” might be more appro_ priate than the term “test amiety’*. Certain ParJlefs between these descriptions and Johnson’s diagnosogenic theory of stuttc:ring are apparent. JoJrnson et al. (1967) described the develop ment of stuttering in the fc’lriwjng way: “AS the child takes on the parent’s anxieties and tensions and I. A to react in terms of them, the normal dis~*=$p’lS fhlencies in his speech tend gradually to become transformed into more ravated and strained hesitations or stoppage reactions”. It is possible that the child who reacts to parental evaluations (as exhibited in their reactions) of his speech disfluency with sufficient intensity to cause that disfluency to persist as stutter@& ma) react simihrrty to parental evaluations of his more general behavior by becoming “test anxious”. Whether some children are more prone than others to WCt strongly to evaluations of their behavior is another aspect of this problem which deserves consideration. A second parallel between test anxiety and stuttering i:: that in both, the can~equence~ [decreased cognitive functioning in the first, severe speech disfluertcy in the second) may persist nlthoqh the initiating conditions are no longer apparent. More thorough investigations are needed of the parent-child relationships of very young children, and more thorough assessments of the children themselves. at the time when their speech dZsfluency begins to concern either them or members of their families. The possibility that the language development of very young stutlerers differs from that af nonstut’terers has been of interest to researchers in recent years. If the results of their investigations demonstrate that any differences which may exist are of sufficient magnitude to influence early academic achievement, the findings reported in this article will have been at least partially explained. Until further research in these areas is carried out, the best hypothesis which can be made regarding the cause of the observed academic retardation in young stutterers, is that it is primarily a function of environmental influences. It is recommended that speech clinicians make it their responsibility to alert the classroom teacher in cases where a stuttering child appears to be functionin;: below hjs capacity. If the teacher’s observations carrespond to those of the clinician, it js possible that modifications can be made in the classroom environment which will help to !mprove the child’s rate of achievement or at least serve to prevent further loss. it is also possible that any improvement in the child’s self-esteem brought about by increased academic success wih in turn foster improvement in his speecJ)Finally, extreme caution should be used in counseling any stutterer and/or his parents with respect to future academic or vocational plans, OI: the basis of his achievement level in the elementary grades, The findings of this investigation do not preclude the possibility of change. It would be extremely wasteful the

of

98

D.E.WlLLIAMS, B.M.MELRQSE and C.L.WCJODS

talents of these children if future plans were established for them on the basis of early achievement test scores, without every effort being made to improve those scores through the establishment of a more favorable academic enviro;?ment. Acknowledgements The work reported herein was performed pursuant to a grant from t?le US Office of Education, Department of Health, Education and Welfare. The authors are grateful to Albert N. Hieronymus and Franklin H. Silverman, University of Iowa, for their assistance in the planning of these investigations. References Buros, O.K., ed., 1959, The fiffh mental meamrements yearbook. Highland Park. NJ., The Gryphon Press. Conradi, E., 1912, Speech defects and intellectual progress. J. educ. Psychol. 3,35 -38. Du eont, M.K., 1946, A comparative study of educational adjustment among stuttering and nonstuttering children. M.A. thesis, University of iowa. Dyer, H., 1961, Is testing a menace to education? N. Y. St. Educ. 49,16-19. Eels, K., 1953, Some implications for school practice of the Chlcago studies of cultural bias in intelligence tests. I&u. Educ. Rev. 23,284-297. Johnson, W., 1959, The onset of stuttering. Minneapolis, University of Minnesota Press. Johnson, W. et al., 1967, Speech handicapped school children. New York, Harper and ROW. Lindquist, E.F. and A.N.Hieronymus, 1964, Manual for administrators,supervisorsand counselors,Iowa Tests of &sic Skills. Boston, Houghton MifflinCo. McDowell, E.D., 1928, Educational and emotional adjustments of stuttering children Teachers College, Columbia University Contributions to Education, No. 314, New York City. Root, A.R., 1926, A survey of speech defectives in the public elementary schools of South Dakota. Elem. Sch. J. 26,53 I-541. Sarason, S. et al., 1960, Anxiety in elementary school children. New York, John Wiley and Sons, Inc. Sarason, S.B., K.T.HiU and PGZimbardo, 1965, A longitudinal study of the relation of test anxiety to performance on intelligence and achievement tests. Monogr. Sot. Res. ChikiDev. 29, no. 7, serial no. 98. Siegel, S., 1956, Nonparametricstatisticsfor the behavioralsciences., New York, McGrawHill. We& R., 1931, The defective in speech. WhireHouse Conference on ChiId Health and ProtecPion. Special Education, F.J.Kelly (chairman). New York, Century Co.