Assessment of French Language Development

Introduction

Language assessment is a complex process. There are several types of tools that can be used to assess children’s language, each with its own advantages and limitations. There are fewer measures available in French than in English, a phenomenon that is partially due to the relative lack of available data on typical French-language development. Several recent efforts, however, have attempted to fill this gap by adapting measures available in other languages or by creating new measures specifically for the French language.

Key questions

How to assess?

Three of the most common approaches to language assessment are 1) formal testing; 2) language samples and observation in a naturalistic context; and 3) parent reports.

Formal tests are the most structured approach to language assessment. The main advantage of tests is that by ensuring that evaluation conditions are the same for each child, the assessor can situate the child in relation to a comparison group (McCauley, 2001). However, the restricted nature of these tests are such that they are rarely used with very young children and even older children can sometimes have weaker performances as a result of a lack of interest or fatigue when the test is long.

Observation and sample collection (e.g. Lund & Duchan, 1993) have the bigger advantage of providing an idea of a child’s performance in a naturalistic context. However, since naturalistic contexts are not very structured, assessors must have a great degree of skill to ensure that they are observing relevant behaviours or can provoke them (Miller & Paul, 1995). Analysis is generally long, given that the child’s outputs must be transcribed for subsequent systematic analysis. Software is available to facilitate the analysis, such as the Systematic Analysis of Language Transcripts (SALT; Miller & Champan, 1998).

Questionnaires for parents often result in a more complete portrait of the child than can be obtained with the other methods (Dale, 1991). Parents can have observed behaviours that are not displayed for the assessor. However, although parents may be genuine experts on the subject of their child, it can be difficult for them to respond to the more pointed questions that the assessor may pose.

How to ensure that the measure is valid?

Regardless of the method used, assessors must ensure that they are using quality tools. They must therefore verify that the measure used was rigorously developed and that it is valid and consistent (for a review, see McCauley, 2001). To be valid, a measure must actually measure what it claims to measure. For example, a measure should be based on solid concepts (construct validity); it should adequately cover the aspects to be assessed (content validity); it should be able to distinguish between those children who are progressing normally from those who are struggling (sensitivity/specificity); it should produce results that are compatible with other comparable measures (concordant validity). To be consistent, the results should truly reflect the child’s capabilities. For example, a child’s performance should be comparable when assessed by two different people (interjudge consistency) or at two different times (test-retest consistency) using the same measure.

No assessment tool is in itself superior to others and the choice of the type of tool will depend on many factors, such as the assessment goal (e.g. detecting potential difficulties vs. measuring progress), the child’s age, and the availability of the tools. This last factor is particularly restrictive for professionals working with Canada’s francophone population (Garcia, Paradis, Sénécal, & Laroche, 2006).

How to overcome the shortage of French-language measures?

Professionals and researchers who want to expand the repertoire of available assessment methods for francophone children in Canada have two main routes: utilize tools currently used with other populations or develop new tools.

Adapting existing tools
The principle of economy suggests that the best approach to take would be adapting existing tools. This course of action may have two major problems. If it is limited to translating a tool (from English, for example) or if an existing tool (in European French, for instance) is simply transferred, the content validity of a tool may be compromised due to the linguistic and cultural differences between the initial population and the Canadian francophone population (Garcia & Desrochers, 1997; Trudeau et al., 1999). A complete adaptation of the measures is therefore called for, to examine in detail the relevance of the content to francophones’ reality in Canada. Canadian clinicians working with francophone populations are aware of the problem and have expressed their dissatisfaction with the cultural relevance, validity and accuracy of available tools (Garcia et al., 2006). Even with a complete adaptation, the results are difficult to interpret if normative data is not collected for the new version (McCauley, 2001). Indeed, anglophone standards cannot be used to interpret the performance of a child tested in French. Here again, francophone clinicians are dissatisfied with the quality of available normative data (Garcia et al., 2006).

Creating new tools
While the adaptation of existing measures is an interesting starting point for certain measures, additional measures are likely necessary for Canadian French. The most important aspects of the language to be measured can be identified while studying the normal development of the French language. Specific assessment tests can be developed once these aspects have been identified. It is also important to better describe the signs of language difficulties in French, in order to identify which behaviours reflect compromised learning. Indeed, given that each language is different, it may be possible that certain types of learning that are specific to French can be more affected by a language difficulty. If this is the case, these types of learning should be measured to detect struggling students (sensitivity/specificity of measures) (Thordardottir, Kehayia, Lessard, Sutton, & Trudeau, 2007). Recent research (Thordardottir & Namazi, 2007) has also suggested that morphological errors (e.g. conjugations) are not as predominant in francophones having a specific language disorder as is observed with anglophones. These authors report a more general grasp of language skills rather than a specific grasp of morphology.

Recent and ongoing research

Several tools that are widely used in anglophone practice have already been adapted. For instance, the Peabody Picture Vocabulary Test, which assesses word comprehension, has been adapted into the Échelle de vocabulaire en images Peabody (Dunn, Thérriault, & Dunn, 1993), which was both standardized and validated. The Test of Auditory Comprehension of Language—Revised, which assesses sentence comprehension, has been adapted into the Épreuve de compréhension du langage Carrow-Woolfolk (Groupe coopératif en orthophonie-Région Laval, Laurentides, Lanaudière, 1995). Finally, the Expressive One-Word Picture Vocabulary Test (EOWPVT), which assesses expressive vocabulary, has also been adapted for Québec. Both of these tests have been standardized for children in kindergarten to second grade. Other work is underway, such as the adaptation of the MacArthur-Bates Communicative Development Inventories, a report for parents of infants aged 8 to 30 months that documents the early phases of communication, into the Inventaires MacArthur du développement de la communication (IMDC; Trudeau, Frank, & Poulin-Dubois, 1999). Data has been collected from almost 1 200 children in Quebec, while a verification of the validity and consistency of the Québécois version has confirmed it as a quality tool. (Trudeau, Boudreault, Marsolais, Rioux, Tousignant, 2005; Boudreault, Cabirol, Trudeau, Poulin-Dubois, & Sutton, 2007). In addition, Elin Thordardottir (2005) has proposed a method for counting the length of phrases in French, and normative data is being collected for children aged 18 months to 6 years. (Sutton & Trudeau, in press; Thordardottir, Kahayia, Sutton, & Trudeau, 2005). This method is compatible SALT software.

The main evaluation tools that have been developed in French and which provide normative data come from Europe. Among these, the best-known are the Épreuves pour l’évaluation du langage (Tests for Language Evaluation) (EPEL; Chevrie-Muller, Simon, & Decant, 1981), and more recently, the Nouvelles épreuves d’évaluation du langage (New Tests for Language Evaluation) (N-EEL; Chevrie-Muller & Plaza, 2001). A recent parent report, the Développement du langage de production en français (Development of Language Production in French) (DLPF; Bassano, Labrell, Champaud, Lemétayer, & Bonnet, 2005) also has potential for the assessment of older children (12-48 months) compared to those targeted by the IMDC. They are not yet widely used in Canada and there is little information available about their validity and consistency. There is currently an effort to standardize certain French-language tools in Quebec. For instance, in a study of 150 children between 18 and 54 months of age, Sutton and team collected data on the performance of children with no language difficulties using the Bulle Test (Auger, 1994; sound production in language) (Poupart, Sutton, Trudeau, Thordardottir, & Lessard, 2006) and the Tours-France Psycholinguistic Laboratory Comprehension Test (Épreuve de compréhension du Laboratoire de psycholinguistique de Tours-France) (Khomsi, 1980; sentence comprehension)(Sutton, Trudeau, Thordardottir, 2006).

Finally, certain tasks currently used in research could become clinical evaluation tools. For example, Royle (2006) has developed a procedure to measure adjective agreement (e.g. vert-verte) in a puzzle format. Similarly, Sutton, Trudeau, Poupart, Grenon, Thordardottir and Lessard (2007) have adapted a task proposed by Nardy (2003) that aims to highlight the liaison phenomenon in French (e.g. les ‘zoiseaux’). These tasks are of interest because they target specific skills in French, and there is no equivalent in the American tools. In the same vein, Béland and Pitchford (2000), and more recently Lefebvre, Girard, Desrosiers, Trudeau and Sutton (2007), have proposed tasks for evaluating phonological awareness (e.g. recognizing rhymes) in young francophones.

Future directions

The next few years will be decisive for French language evaluation in Quebec. Indeed, several of the projects mentioned will be coming to a close, and making available normative data. There is still much to do, however, particularly with francophone populations outside of Quebec. In fact, just as certain cultural or linguistic differences make it difficult to directly transfer European tools to the Canadian population, it would not be appropriate to attempt to generalize Québécois tools and standards to minority francophone populations. As well, the country’s changing cultural reality is such that an increasing number of children are growing up in a bilingual, or multilingual, milieu. In 2001, almost one in six Canadians (more than 5 000 000 people) reported that they spoke English and French (Statistics Canada, 2007a). Further, one out of every six Canadians spoke neither English nor French as their first language (Statistics Canada, 2007b). It will therefore be necessary to develop more accurate assessment methods for these children, taking into account new understanding of the development of bilingual children.

Conclusion

Developing appropriate methods is a complex and long-winded process. French-language evaluation poses its own challenges and is part of a unique Canadian context. However, by basing themselves on solid developmental data and by taking a rigorous approach, French-speaking professionals should have nothing preventing them from having access to quality methods to assist them in their work and to offer children the services they need.

References

Auger, D. (1994). Casse-tête d’évaluation de la phonologie. Montréal: Author.

Bassano, D., Labrell, F., Champaud, C., Lemétayer, F., & Bonnet, P. (2005). Le DLPF: un nouvel outil pour l’évaluation du développement du langage de production en français. Enfance, 57, 171-208.

Boudreault, M.-C., Cabirol, É.-A., Trudeau, N., Poulin-Dubois, D., & Sutton, A. (2007). Les Inventaires Macarthur du développement de la communication : validité et données normatives préliminaires. Journal of Speech-Language Pathology and Audiology, 31(1), 27-37.

Chevrie-Muller, C., Simon, A. M., & Decante, P. (1981). Epreuves pour l'examen du langage (EPEL). Paris: Editions du Centre de Psychologie Appliquée.

Chevrié-Muller, C., & Plaza, M. (2001). Nouvelles Épreuves d’évaluation du langage. Paris: Editions du Centre de Psychologie Appliquée.

Courcy, A., Beland, R., & Pitchford, N. J. (2000). Phonological awareness in French-speaking children at risk for reading disabilities. Brain & Cognition, 43(1-3), 124-130.

Dale, P. S. (1991). The validity of a parent report measure of vocabulary and syntax at 24 months. Journal of Speech and Hearing Research, 34, 565-671.

Dunn, L., Thériault-Whalen, C., & Dunn, L. M. (1993). Échelle de vocabulaire en images Peabody. Toronto: Psycan.

Frank, I., Poulin-Dubois, D., & Trudeau, N. (1997). Inventaire MacArthur de la communication: Mots et énoncés. [Québécois adapatation of the MacArthur Communicative Development Inventory: Words and Sentences].

Garcia, L. J, & Desrochers, A. (1997). L’évaluation des troubles du langage et de la parole chez l’adulte francophone. La Revue d’orthophonie et d’audiologie, 21, 271-293.

Garcia, L. J., Paradis, J., Sénécal, I., & Laroche, C. (2006). Utilisation et satisfaction à l'égard des outils en français évaluant les troubles de la communication / Use of and satisfaction with assessment tools for evaluating communication disorders in French. Journal of Speech-Language Pathology and Audiology, 30(4), 239-249.

Groupe coopératif en orthophonie-Région Laval, Laurentides, Lanaudière (1995) Épreuve de compréhension de Carrow-Woolfolk et Expressive One-Word Vocabulary Test, French adaptation.

Lefebvre, P., Girard, C., Desrosiers, K., Trudeau, N., & Sutton, A. (2007) Phonological awareness tasks for French-speaking preschoolers, manuscript submitted for publication.

Lund, N. J., & Duchan, J. F. (1993). Assessing children’s language in naturalistic contexts, 3^rd edition. Toronto: Prentice-Hall Canada.

McCauley, R. J. (2001). Assessment of language disorders in children. Mahwah, NJ: Lawrence Erlbaum.

Miller, J., & Chapman, R. (1984–2002). Systematic Analysis of Language Transcripts: Software for Analyzing English and Spanish Language Samples. Madison, WI: Language Analysis Laboratory, University of Wisconsin-Madison.

Miller J., & Paul, R. (1995). The clinical assessment of language comprehension. Baltimore, MD: Paul H. Brookes.

Nardy, A. (2003). Production et jugement d’acceptabilité entre 2 et 6 ans : Aspects psycholinguistiques et sociolinguistiques de l’acquisition des liaisons. Mémoire de recherche. Laboratoire de linguistique et didactique des langues étrangères et maternelles, Université Stendhal, Grenoble.

Poupart, A., Sutton, A., Trudeau, N., Thordardottir, E., Lessard, N. (2006, June). Évolution de la phonologie chez les enfants de 24 à 54 mois. CLLRNet conference, Charlottetown, PEI.

Royle, P. (2006). Le casse-tête comme outil de recherche, de dépistage, et d’intervention en dysphasie. Fréquences, 18, 36-37.

Statistics Canada (2007a). Retrieved July 11, 2007 from
http://www12.statcan.ca/english/census01/products/analytic/companion/lang/canada.cfm#%20multilingual.

Statistics Canada (2007b). Retrieved July 11, 2007 from
http://www12.statcan.ca/english/census01/products/analytic/companion/lang/bilingual.cfm#%20increase.

Sutton, A., & Trudeau, N. (in press). Nos bouts de choux : linguistes en herbes. Médecine Sciences, 11 p.

Sutton, A., Trudeau, N, Poupart, A., Grenon, V., Thordardottir, E., & Lessard, N. (2007, April) Développement phonologique chez les enfants francophones âgés de 18 à 54 mois. Canadian Association of Speech-Language Pathologists and Audiologists (CASLPA) conference, Moncton, NB.

Sutton, A., Trudeau, N., & Thordardottir, E. (2006, June). Compréhension lexicale, grammaticale et pragmatique chez les enfants québécois d’âge préscolaire. Congrès de l’Ordre des orthophonistes et audiologistes du Québec, Gatineau, QC.

Thordardottir, E. (2005). Early lexical and syntactic development in Quebec French and English: Implications for cross-linguistic and bilingual assessment. International Journal of Communication Disorders, 40, 243-278.

Thordardottir, E., Kehayia, E., Lessard, N., Sutton, A., & Trudeau, N. (2007). Diagnostic accuracy of French language measures for the identification of SLI at age 5. Society for Research in Child Language Development, Madison, WI, June 7-9.

Thordardottir, E., Kahayia, E., Sutton, A., & Trudeau, N. (November 2005). Language sample measures for French-speaking 5-year-olds. American Speech-Language-Hearing Association, San Diego, CA.

Thordardottir, E., & Namazi, M. (2007) Specific language impairment in French-speaking children: Beyond grammatical morphology. Journal of Speech, Language, and Hearing Research,50(3), 698-715.

Trudeau, N., Boudreault, M.-C., Marsolais, I., Rioux, A., Tousignant, J. (2005, June). Validité et fidélité de la version québécoise de l'Inventaire MacArthur du développement de la communication. CLLRNet conference, Toronto, ON.

Trudeau, N., Frank, I., & Poulin-Dubois, D. (1999) Une adaptation en français québécois du MacArthur Communicative Development Inventory. La revue d’orthophonie et d’audiologie,23, 61-73.