Notes on language experience

Notes on Language Experience    Stephen Krashen

August, 2016  DRAFT

Language Experience was frequently used and discussed several decades ago, but is not mentioned much these days, possibility because there is no chance for publishers to make money from it.   It was popular in first grade classes for native speakers of English in the United States, but it can easily be used in first or second language development, for beginners up to low intermediate of any age, and could be a powerful source of written comprehensible input.

There are several versions of language experience, but they all are in agreement with Hall's definition (Hall, 1978): "... a method in which instruction is built on the use of reading materials created by writing down children's spoken language." (p. 2).¹

Here is one simple manifestation:

1.     The student dictates a very short story or anecdote to the teacher.

2.     The teacher writes out the story or anecdote (these days using a word processor) and makes copies of the story.

3.     The student and other students read the story, which could become part of the classroom library.

Current research and theory predicts that language experience will work: It is compelling comprehensible input,  and fully utilizes one of the most powerful ways of making input compelling: personalization. The stories children dictate are usually about themselves, their interests, and their lives. Language experience thus has a strong similarity to TPRS (Teaching Proficiency through Reading and Story Telling), a method of teaching second languages that relies on stories co-constructed by teachers and students (Ray and Seely, 2015). Studies thus far show that TPRS is effective for second language acquisition (Dziedzic, J. 2012; Varguez, K. 2009. Watson, 2009; Pippins and Krashen, 2015).

I present here a brief review of language experience research.  None of the studies used exclusively language experience.  In every case, it was mixed with

direct instruction (e.g. word attack skills, phonics).  Comparison groups always had direct instruction as well, so this is probably not a serious confound. Except for McCanne (1966), all  studies in table 1 deal with English as a first language.

Table 1 presents only grade one²  results for "Paragraph Reading" tests in which children have to understand what they read. The results presented here agree closely with the Bond and Dykstra (1967) presentation of the Hahn, Kendrick, Stauffer and Vilscek  et. al. (reported as Cleland, p. 127-8) studies , in which means were adjusted for pretest performance on measures of word reading, vocabulary, spelling, letter names, IQ and other tests (table 46, p. 75).

 Table 1: Research on the Effectiveness of Language Experience, comparied to Basal Programs (positive = language experience superior to traditional (basal) group).

Study	duration	pos	nd	neg	N (exp/comp)	ESc
journal papers
Vilscik, Morgan & Cleland, 1966	1 year	X				no data
Harris and Serwer, 1966	1 year			xd	exp = 1,141	no data
Stauffer, 1966	1 year	X			232/201	no data
Kendrik & Bennett, 1966	1 year			x	645/657	-0.41
McCanne, 1966	1 year			xe	98/98/98	no data
Hahn, 1966	1 year	xa,b			55/55	0.13
dissertations
Hall, 1965	1 semester	X
M. Stauffer	1 year	X
grant report
Lamb, 1971	1 year		x

(a)   A third group did language experience as well, but also included instruction in the International Teaching Alphabet.  The experimental group did slightly better than the ITA group (d=.13) and the ITA group outperformed the basal comparison group (d = .13).  The Language Experience group did about twice as much self-selected reading than the other groups, reading an average of 15.5 books in a month in grade 2.

(b)   Significance fell just short of the .05 level, one-tail, with t = 1.60 (df = 108), p = .056. The difference was easily significant at the .10 level.  Calculation by SK.

(c)   Effect sizes (ES) were calculated from data presented in journal papers, but this was only possible in two cases.

(d)   Harris and Serwer point out that the basal reader method "held a slight lead ... It was associated with slightly but significantly highest results in meaningful silent reading comprehension" (p. 634). Unfortunately, it was not possible to calcualte the effect size. 

(e)   McCanne reported that "reading comprehension" included two subtests, but did not describe the tests.

Table 2: Research summary

	positive	no dff	negative
Dissertations	2
Journal papers	3		3
grant report		1
Total	5	1	3

The scorecard presented in table 2 shows a slight advantage for Language Experience, with the advantage coming from dissertation research. Dissertation research is sometimes thought to be lower quality than published research, but Glass, McGaw and Smith (1981) concluded that dissertation reearch is of slightly higher quality in terms of design than published research (p. 51). Also, experimental effects reported in dissertations are typically lower than effects reported in published studies, reporting less support for favored hypotheses (p. 67, 226).

Note also that the advantage for the basal group in Harris and Serwer (1966), while significant, was "slight" (comment d, table 1) and the results of only one study are firmly negative (Kendrick and Bennett, 1966).

This is a crude analysis: There was a wide variation in procedure among the studies (see Vilscek, 1968) and details of several studies were not available to me. The only commonality is that reading is largely done from texts dictated by the students. 

What is clear, however, is that language experience deserves another look, under more carefully controlled experimental conditions.  Language experience  stories have tremendous potential: They are free, and can form the basis for a classroom library, and can be easily shared on the internet.

Notes

(1) Some history: "The story (of language experience) begins over sixty years ago [this article was published in 1965] when Miss Flora J. Cooke, a teacher at the Chicago Institute, later at the Francis Parker school, Chicago, began experimenting with a 'natural' method of teaching beginners to read through recording on the blackboard the children's oral expressions relating to current experiences. Miss Cooke stated her hypothesis concerning this innovation. Children may learn to read as naturally as they learn to talk, she said ... She found that the children readily learned to read records she prepared of their own experiences without following any particular method of teaching reading." ( p. 280). (From Hildreth, 1965).

(2) Language experience for older children (grades 2 and 3) sometimes included self-selected reading (Harris, Serwer, and Gold, 1967).  Such a study is no longer a test of language experience alone, but is still a test of the more general Comprehension Hypothesis.  It should be pointed out, however, that regular book reading happened in grade 1 studies as well. In Vilscek, Morgan, and Cleland (1966), a first grade study, children "transitioned" to "instructional reading in trade books" when each child reacher the "primer instructional level" (p. 35). Another first grade study, Kendrick and Bennett (1966) included reading library books and stories from textbooks as "reading activities"  as well as language experience (p. 95).

References

Bond, G. and Dykstra, R. 1976. The cooperative research program in first-grade reading instruction. Reading Research Quarterly 2(4) (Entire issue).

Dziedzic, J. 2012. A comparison of TPRS and traditional instruction, both with SSR. International Journal of Foreign Language Teaching 7(2): 4-6.

Glass, G., McGaw, B., and Smith, M.L. 1981. Meta-analysis in Social Research. Beverly Hills: Sage Publications.

Hahn, H. 1966. Three Aproaches to Beginning Reading Instruction - ITA, Language Experience and Basal Readers. The Reading Teacher 19(1966) 590-594.

Hahn, H. 1967.  Three approaches to beginning reading instrution – ITA, language experience, and basal readers – extended to second grade.  The Reading Teacher 26: 711-715.

Harris, A. and Morrison, C. 1969. The CRAFT project: A final report. The Reading Teacher 22: 335-340.

Harris, A. and Serwer, B. 1966. The CRAFT project: instructional time in reading research. Reading Research Quarterly 2: 29-56.

Harris, A. and Serwer, B. 1966. Comparing reading approaches in first-grade teaching with disadvantaged children. The Reading Teacher 19: 631-635.

Harris, B., Serwer, B. and Gold, L. 1967. Comparing reading approaches in first grade teaching with disadvantaged children = extended into the second grade. The Reading Teacher 20: 699-703.

Hildreth, G. 1965. Experience-related reading for school beginners. Elementary English 42: 280-292.

Kendrick, W. 1966. A comparative study of two first grade language arts programs. The Reading Teacher 20: 23-30.

Mallett, G. 1977. Using language experience with junior high native Indian students. Journal of Reading 21: 25-28.

McCanne, R. 1966. Approaches to first grade English reading instruction for children from Spanish-speaking homes. The Reading Teacher 70:70-75.

Pippins, D. and Krashen, S. 2016. How well do TPRS students do on the AP? International Journal of Foreign Languge Teaching. 11(1): 25-30.

Ray, B. and Seely, C. 2015. Fluency Through TPRS Storytelling, Expanded 7th Edition. Berkeley: Command Performance Institute.

Vilscek, E. 1968. What research has shown about the language-experience program. In Vilscek, E. (Ed.), A Decade of Innovations: Approaches to Beginning Reading. Proceedings of the Twelfth Annual Convention, International Reading Association. Newark, Delaware: International Reading Association. pp. 9-23.

Vilscek, E., Morgan, L. and Cleland, D. 1966. Coordination and integrating language arts instruction in first grade.  The Reading Teacher 20:31-37.

Varguez, K. 2009. Traditional and TPR Storytelling instruction in the Beginning High School Spanish Classroom. International Journal of Foreign Language Teaching 5 (1): 2-11.

Watson, B. 2009. A comparison of TPRS and traditional foreign language instruction at the high school level. International Journal of Foreign Language Teaching 5 (1): 21-24.