Reference Work Entry

Encyclopedia of Autism Spectrum Disorders

pp 2903-2915

Social Language Development Test

  • Linda BowersAffiliated withLinguiSystems, Inc Email author 
  • , Rosemary HuisinghAffiliated withLinguiSystems, Inc

Synonyms

SLDT-A; SLDT-E; Social language development test-adolescent; Social language development test-elementary

Description

The tests of social language development-elementary and adolescent (Bowers, Huisingh, & LoGiudice, 2008, 2010) are diagnostic tests of social language skills. They are designed to determine the role language development plays in the acquisition of social understanding for students from the ages of 6.0–17.11 years from the following educational settings: regular education (no active IEP) and special education (active IEP). Students with a diagnosis of autism and delayed language development were included in the item pool and standardization studies.

The testing materials are as follows:

The social language development test-elementary (SLDT-E) includes an examiner’s manual, the scoring standards and example responses book, a picture stimuli book, and test forms. Ages include 6.0–11.11 years.

The SLDT-E is comprised of four subtests designed to differentiate how students with language impairments and autism differ from their typically developing peers in social cognitive processing, in identifying feelings of participants in a conflict, in identifying and evaluating strategies to overcome obstacles, and in knowing when a conflict is resolved. The subtests are the following: (a) making inferences, (b) interpersonal negotiation, (c) multiple interpretations, and (d) supporting peers. Each subtest has 12 items.

Many children with language impairments (LI) exhibit poor social interaction with others, and these social differences may appear as early as preschool and continue or intensify as these children mature (Brinton, Robinson, & Fujiki, 2004; Cohen et al., 1998; Craig, 1993; Fujiki, Brinton, Robinson, & Watson, 1997; Hadley & Rice, 1991). A tool that would pinpoint areas of strengths and weaknesses would guide speech-language pathologists (SLPs) in determining appropriate intervention goals and objectives. Tasks in this test focus on taking someone else’s perspective, making correct inferences, negotiation conflicts with peers, being flexible in interpreting situations, and supporting friends diplomatically.

The social language development test-adolescent (SLDT-A) includes an examiner’s manual, the scoring standards and example responses book, a picture stimuli book, test forms, and an audio CD used with subtest E: interpreting ironic statements. Ages include 12.0–17.11 years.

The SLDT-A is comprised of five subtests designed to differentiate typically developing elementary children from typically developing adolescents, typically developing adolescents from language-impaired adolescents, and language-impaired adolescents from adolescents identified with high-functioning autism. The subtests are the following: (a) making inferences, (b) interpreting social language, (c) problem solving (stating and justifying solutions), (d) social interaction, and (e) interpreting ironic statements. Each subtest has 12 items each. Three of the subtests have two questions per item.

Language is the skill that is central to engaging in all aspects of peer relationships. Students with limited language skills are susceptible to misinterpretations and have more difficulty participating in the highly verbal, rapidly delivered social processes used in peer group formation (Gallagher, 1993). They often have difficulty maintaining friendships, negotiating conflict, and getting along with others. Tasks in this test focus on taking someone else’s perspective, making correct inferences, solving problems with peers, interpreting social language, and understanding idioms, irony, and sarcasm.

These tests do not address all aspects of social language or pragmatic skills. Rather, they focus on social interpretation and interaction with peers and/or friends. For both tests, the items are not arranged in order of difficulty because, as of yet, no hierarchy is established for these tasks. For this reason, there are no basals or ceilings; each subtest is administered in its entirety to every student.

The tests begin with the first item in subtest A. Explicit directions for how to administer the items are provided. Demonstration items are provided for each subtest. These items may be altered or explained to show the student how to respond.

Following the demonstration item, each subtest begins with item 1 regardless of the student’s age. Each item is presented verbally. The only exception is subtest E: interpreting ironic statements on the SLDT-A. An audio CD is used to administer this subtest.

Allowable prompts, used only as worded, are listed on the test forms. These are used if the student’s response is unclear. These prompts are not used to give the student “a second chance” after a clear, complete, but incorrect response.

More than one testing session is allowable if the break between sessions occurs upon completion of a subtest and all of its questions.

The test examiner must be a trained professional familiar with language disorders (e.g., speech-language pathologist, psychologist) as they require careful interpretation of responses by the examiner.

Subtests A in both tests tap the student’s ability to infer what someone in a photo is thinking. The student also states the visual clues that facilitated making the inference. Skills assessed in these subtests include the ability to:
  • Detect nonverbal and context clues in a picture of a person or people

  • Assume the perspective of a specific person in a picture

  • Infer what the person is thinking

  • Express the person’s thought as a relevant, direct quotation

  • State the visual clue that suggests what the person is thinking

A score of 1 or 0 is assigned to each response based on the relevancy and quality of the response. For the SLDT-A, the student must provide an appropriate response for both questions to earn a score of 1.

SLDT-E Subtest B: Interpersonal Negotiation asks the student to imagine being involved in a given conflict with a friend. The student states the problem (task a), proposes an appropriate solution (task b), and explains why that would be a good solution (task c). No pictures or photos are used for this subtest. Skills assessed in this subtest include the ability to:
  • Understand a short passage about a conflict with a friend

  • Infer the perspective of each person in the conflict

  • State the problem clearly

  • Propose a solution

  • Explain why that would be a good solution

A score of 3, 2, 1, or 0 is assigned to each response based on relevancy and quality.

SLDT-A Subtest B: Interpreting Social Language asks the student questions about how people communicate. The student is asked to give an example and an appropriate context in which the type of communication would be used. Skills assessed include the ability to:
  • Demonstrate actions

  • Tell an appropriate reason or use for an action

  • Think/talk about language

  • Interpret figurative language including idioms

A score of 1 or 0 is assigned to each response based on relevancy and quality.

SLDT-E Subtest C: Multiple Interpretations asks the student to provide two distinctively different, plausible interpretations of the same photo. Skills assessed in this subtest include the ability to:
  • Logically interpret a situation in a photo in two different ways

A score of 1 or 0 is assigned to each response, based on relevancy and quality.

SLDT-A Subtest C: Problem Solving (Stating and Justifying Solutions) asks the student to solve a problem by stating and justifying a logical solution. The student must give both a solution and a justification. Skills assessed include the ability to:
  • State an appropriate solution to a problem in a situation with a peer

  • Justify the solution

  • Negotiate conflicts with peers

  • Take the perspective of the other person in the situation

  • State the conflict from a mutual perspective

A score of 1 or 0 is assigned to each response based on relevancy and quality.

SLDT-E Subtest D: Supporting Peers asks the student to take the perspective of someone involved in a situation with a friend. The student tells what to say in reaction to the friend’s situation. Responses receive credit based on the degree of support they offer to the friend, not on the truthfulness.

Being truthful is a basic maxim of interpersonal communication (Grice, 1980), yet speakers are also expected to help, not hurt, their communicative partners (Lakeoff, 1973; Sweetser, 1987). Telling white lies successfully requires reconciling these apparently contradictory communication rules and knowing when and how to adapt them to suit various social situations (Talwar, Murphy, & Lee, 2007).

A score of 4, 3, 2, 1, or 0 is assigned to each response based on relevancy and quality of support for a friend.

SLDT-A Subtest D: Social Interaction asks the student to listen to situations and answer questions about them. In some situations, an appropriate response may include a dishonest or rude remark. Skills assessed include the ability to:
  • Listen to short passages

  • Understand social interactions with peers

  • Provide an appropriate, supportive response

  • Ignore the situation (when doing nothing is the best option)

A score of 1 or 0 is assigned to each response based on the relevancy and quality of support for the situation.

SLDT-A Subtest E: Interpreting Ironic Statements asks the student to listen to some situations on a CD. The narrator reads the situations and asks what someone means at the end of each one. Research on intonation features associated with irony is inconsistent, and ironic speech does not necessarily include “signature acoustic features” (Bryant & Fox Tree, 2005; Nakassis & Snedeker, 2002). For this reason, the assumed features of irony (e.g., exaggerated pitch) were “toned down” so that the student must rely on context discrepancies and other context clues to determine the ironic meaning of a statement. Our research proved that including exaggerated intonations made the task too easy for the age range of this test. Skills assessed included the ability to:
  • Understand common idioms used in everyday language

  • Reject the literal meaning of the statement

  • Understand the speaker’s belief

  • Judge the speaker’s attitude

  • Recognize sarcasm and interpret irony

A score of 1 or 0 is assigned to each response based on the relevancy and quality.

Historical Background

In 1983, we published a test of pragmatics called the interpersonal language skills assessment (ILSA) by speech-language pathologists, Carolyn Blagden (now LoGiudice) and Nancy McConnell. This test of pragmatic behaviors looked at interaction of 8–13-year-olds while playing a board game. The authors wanted to determine if there was age progression and reverse age progression on such behaviors as advising, commanding, accusing, deprecating, justifying, and supporting. They found that in general, as students got older, their outright negative comments decreased while their sarcasm increased. They also found out that comments with grammar or semantic errors decreased with age. Although some speech pathologists were curious about the skills assessed in the ILSA, there was a general lack of interest from the field. By 1989, this test was out of print.

Between 1990 and 2008, we continued to scour the research on pragmatic/social language skills. Our hypothesis was that children develop social “governors” as they mature. Over time, research articles about peer support, negotiating, making inferences, and interpreting facial expression and body language appeared in the literature. We also read research about the development of these skills in children on the autism spectrum and on children who had language disorders but who were not classified as autistic.

Despite a growing body of research, as of yet, there is no well-recognized, research-based developmental model for children learning social interaction skills. That may be because the following:
  • Adults seem to have mastered social interaction without any conscious effort or formal instruction.

  • Much of social language is abstract and covert and cannot be observed directly. We can only assume what someone else is thinking based on nonverbal cues, background knowledge, and the person’s words and actions.

We do know that children with language impairments exhibit poor social interaction with others. We decided that if speech pathologists had an effective, standardized tool to assess the social interaction skills of students that would also pinpoint areas of strengths and weaknesses, it would guide the SLP in qualifying students for therapy and help in determining appropriate intervention goals and objectives.

We began writing and rewriting items and testing and retesting children. Our item pool results showed us the need for a scoring system that was sensitive to subtle changes in verbal behavior. The standardization studies proved the scoring standards were correct but that further refining was necessary. These tests represent the important data we gathered and we believe are a good start toward identifying children with social language differences.

Psychometric Data

Item pool and standardization studies were conducted for both tests. The items and subtests for both tests were subjected to rigorous statistical analyses in these studies.

The item pool and standardization studies for each test included subjects from regular education, special education, all socioeconomic levels, and from various racial groups, including White, Black, Hispanic or Latino, and other mixed racial group. Subjects with IEPs for special services but who attended regular education classes were also included. Subjects who did not use English proficiently at school, were nonverbal, had any degree of hearing loss, or resided outside the United States were excluded from the standardization studies for each test. Subjects were from all geographic regions of the United States. The sample population reflected the national school population demographics from the 2004 national census for race, gender, age, and educational placement for the item pool and standardization studies.

For both tests, the item pool items were administered to random samples of subjects at yearly age intervals. Subjects included in the item pool studies were not included in the standardization studies.

A team of expert speech-language pathologists determined credit levels for item pool responses during a response analysis session. Items that did not show statistical age progression, showed bias, or were not deemed fair for an ethnic group were eliminated. Item difficulty indexes and item discrimination indexes were computed for each item at each of the yearly age levels. Items retained for the final versions of the test met two empirical criteria:
  1. 1.

    Demonstrate age progression in terms of increasing percents of subjects passing at successive age levels

     
  2. 2.

    Demonstrate significant discrimination between high and low scorers on the subtest at each age level

     

The test examiners were speech-language pathologists who held a master’s degree or higher from an accredited university.

SLDT-E

Item Pool

  1. 1.

    Four subtests, 68 items.

     
  2. 2.

    N = 390 subjects ages 6 years, 0 months to 11 years, 11 months.

     
  3. 3.

    Data was gathered at yearly age intervals.

     

Standardization

  1. 1.

    Four subtests, 48 items.

     
  2. 2.

    N = 1,104 subjects ages 6 years, 0 months to 11 years, 11 months.

     
  3. 3.

    Data was gathered at half-year age intervals. The item pool analysis showed that significant changes in social language development could be shown by grouping subjects in this manner.

     
  4. 4.

    Test examiners were required to make value judgments regarding the appropriateness of responses and to score them as indicated on the scoring standards provided to them.

     
  5. 5.

    The final version of the test has four subtests, 12 items per subtest for a total of 48 items.

     

SLDT-A

Item Pool

  1. 1.

    Five subtests, 89 items.

     
  2. 2.

    N = 500 subjects ages 12 years, 0 months, to 17 years, 11 months.

     
  3. 3.

    Data was gathered at yearly age intervals.

     

Standardization

  1. 1.

    Five subtests, 69 items.

     
  2. 2.

    N = 834 subjects ages 12 years, 0 months, to 17 years, 11 months.

     
  3. 3.

    Data was gathered at half-year age intervals. The item pool analysis showed that significant changes in social language development could be shown by grouping subjects in this manner.

     
  4. 4.

    Test examiners were required to make value judgments regarding the appropriateness of responses and to score them as indicated on the scoring standards provided to them.

     
  5. 5.

    The final version of the test has five subsets, 12 items per subtest for a total of a 60 items.

     

 Mean and median raw score values and standard deviations for each subtest and the total test score were computed for both tests. There were no significant gender differences which supported the use of combined male-female norms for the elementary and adolescent tests.

For reporting purposes, three types of scores are used: age equivalents, percentile ranks, and standard scores.

Reliability was established by both the use of test-retest and internal consistency methods. Table 1 represents the internal consistency results for the SLDT-E, and Table 2 represents the internal consistency results for the SLDT-A.
Social Language Development Test, Table 1

Test-retest reliability coefficients and standard errors of measurement for each task, each subtest, and total test by age

Chronological age

N

Making inferences

Interpersonal negotiation

Multiple interpretations

Supporting peers

Total test

Task a

Task b

Making inferences total

Task a

Task b

Task c

Interpersonal negotiation total

r

SEM

r

SEM

r

SEM

r

SEM

r

SEM

r

SEM

r

SEM

r

SEM

r

SEM

r

SEM

6-0 … 6-5

9

.94

0.79

.46

2.35

.87

1.98

.64

6.12

.25

6.30

.40

6.82

.59

15.21

.70

1.73

.66

5.86

.59

23.11

6-6 … 6-11

3

.50

2.46

1.00

0.00

.99

0.66

−.40

12.24

1.00

0.26

.99

0.82

1.00

1.16

.98

0.43

.99

0.95

1.00

1.44

7-0 … 7-5

3

.97

0.52

.94

0.71

.98

0.81

−.20

7.76

.80

9.05

1.00

0.40

.94

4.74

.93

0.76

.97

1.87

.98

4.81

7-6 … 7-11

11

.75

1.63

.42

2.22

.67

3.02

.87

2.50

.70

2.95

.72

3.59

.81

7.19

.59

1.90

.81

4.02

.91

8.64

8-0 … 8-5

12

.91

1.09

.86

1.16

.91

1.70

.80

2.78

.91

1.74

.84

2.65

.87

5.89

.53

1.91

.86

3.19

.88

9.75

8-6 … 8-11

8

.14

3.08

.53

1.85

.58

3.44

.66

4.11

.82

2.93

.86

2.59

.92

5.06

.59

1.84

.65

5.40

.76

14.54

9-0 … 9-5

7

.83

1.22

.74

1.01

.87

1.58

.57

4.28

.72

2.89

.94

1.43

.80

6.55

.75

1.41

.92

2.41

.90

7.69

9-6 … 9-11

8

.67

1.67

.77

0.97

.83

1.75

.05

4.86

.63

2.98

.69

3.10

.56

8.48

.78

1.30

.86

2.95

.66

12.64

10-0 … 10-5

6

.44

2.08

.82

1.02

.72

2.53

.90

1.94

.28

4.87

.05

5.92

.12

14.89

.67

1.59

.53

5.06

.53

17.21

10-6 … 10-11

9

.36

2.47

.69

1.03

.48

3.03

.48

2.69

.80

2.07

.28

4.54

.57

6.88

.86

0.99

.58

4.30

.58

11.34

11-0 … 11-5

10

.72

1.82

−.30

2.99

.58

3.53

.39

5.56

.68

3.81

.53

4.98

.84

7.77

.88

1.04

.89

2.77

.78

15.27

11-6 … 11-11

10

.35

2.35

.88

0.70

.70

2.32

.69

2.53

.69

3.20

.90

1.94

.94

3.35

.57

1.67

.30

6.75

.87

8.71

Median

 

.63

1.77

.65

1.33

.77

2.20

.45

4.78

.56

3.59

.68

3.23

.75

7.26

.74

1.38

.75

3.79

.79

11.26

r reliability coefficient, SEM standard error of measurement

Social Language Development Test, Table 2

Reliability based on item homogeneity: Kuder-Richardson (KR20) coefficients for each task, each subtest, and total test by age

Chronological age

Making inferences

Interpersonal negotiation subtest

  

Total test

Task a

Task b

Making inferences total

Task a

Task b

Task c

Interpersonal negotiation total

Multiple interpretations

Supporting peers

6-0 … 6-5

.75

.73

.84

.84

.76

.83

.93

.74

.77

.95

6-6 … 6-11

.78

.76

.86

.86

.77

.83

.93

.71

.77

.96

7-0 … 7-5

.70

.74

.82

.75

.71

.78

.89

.68

.76

.94

7-6 … 7-11

.75

.73

.82

.81

.67

.75

.90

.70

.74

.93

8-0 … 8-5

.78

.77

.86

.80

.66

.75

.88

.65

.66

.93

8-6 … 8-11

.76

.74

.84

.83

.73

.78

.91

.68

.77

.95

9-0 … 9-5

.71

.62

.80

.83

.52

.75

.87

.68

.77

.93

9-6 … 9-11

.70

.63

.79

.76

.56

.68

.85

.67

.73

.92

10-0 … 10-5

.69

.74

.83

.78

.70

.76

.90

.67

.62

.93

10-6 … 10-11

.76

.60

.80

.38

.46

.69

.77

.65

.57

.87

11-0 … 11-5

.80

.78

.88

.89

.78

.82

.94

.74

.83

.97

11-6 … 11-11

.73

.67

.80

.45

.44

.72

.75

.65

.65

.89

Median KR20

.74

.71

.83

.75

.65

.76

.88

.68

.72

.93

Empirical validity for each test was established with contrasted groups. Validity was established by comparing the test performances of randomly selected subjects from the sample population with a matched sample of subjects with language disorders and autism spectrum disorders receiving special services. The SLDT-E and the SLDT-A significantly discriminate between these clinical groups at most age levels. The exception is on the SLDT-E for ages 6–0 through 6–5. In this age group, the normative sample and the language-disordered group did not show a significant difference for the total test score.

Construct validity was established with point biserial correlations. Inspection of these correlations revealed highly satisfactory levels of item consistency for both tests (88% for the SLDT-E and 97% for the SLDT-A). The subtest intercorrelations and the correlations of individual subtests with the total test suggest that the subtests do assess separate social language functions but also measure a common general dimension. Overall, internal consistency estimates were satisfactory.

Race/socioeconomic group differences were conducted at the subtest level for both tests. Random samples of White, Black, and Hispanic or Latino student performances at the item level were compared by analyzing the proportion of students passing each item and at the subtest level by:
  • Chi-square analyses to determine if significant relations existed between race and test performance

  • Two-factor analysis of variance (ANOVA) analyses

In the SLDT-E, there were race differences in only 11% of the total number of the possible statistical tests for race and socioeconomic status (SES) group differences in only 18% of the statistical tests for that factor. In the SLDT-A, there were race and SES group differences in only 5% of the total number of the possible statistical tests for race and SES differences in only 15% of the statistical tests for that factor.

For both tests, the differences were usually small with most of them in the magnitude of one or two raw scores. Collectively, the results indicate that neither race nor SES group has a major impact on possible race or SES group bias for either the elementary or adolescent test. Further study is warranted however.

Clinical Uses

The results obtained from diagnostic testing should help the examiner to:
  1. 1.

    Identify the student’s strengths and weaknesses

     
  2. 2.

    Make recommendations for additional testing

     
  3. 3.

    Make well-founded, educationally significant recommendations for remediation

     
  4. 4.

    Help teachers, parents, guardians, and the student understand the nature of the student’s social language functioning

     
  5. 5.

    Make well-founded recommendations for boosting social success in school and at home

     

A student’s performance on either of these tests may relate to his academic performance and peer interaction. A student may perform poorly on these tests because he has below average vocabulary skills, difficulty recognizing and interpreting facial expressions and body language, does not know the steps involved in processing a photo of someone, etc.

Performance on the individual subtests and total test performance, along with additional diagnostic tests and parent/teacher report, will give the SLP a framework for developing a remediation program that addresses the student’s social language weaknesses and builds on the student’s strengths.

For each subtest on the SLDT-E and SLDT-A, the authors identified error patterns and, based on those, identified remediation strategies for the SLP. These are described in the examiner’s manuals.

As previously stated, these tests do not address all aspects of social language or pragmatic skills. They focus on social interpretation and interaction with peers and/or friends and are the first tests to have normative data on which to base therapeutic decisions. The authors encourage further research into social language skill development.

See Also

Asperger Syndrome

High-Functioning Autism (HFA)

Norm-Referenced Testing

Copyright information

© Springer Science+Business Media New York 2013
Show all