On other pages part 1, part 2, part 4, part 5.


[See newsletter items by George O'Halloran.]


George O'Halloran.

This inventory took its origin from a device I used during the war years in West Africa to allocate African soldiers for training. It was easy enough to assign those with recognised certificates in English or those who were completely illiterate but in between those two groups there was a number who were literate in varying degrees. To fit these men effectively into training schemes it was necessary to find out reasonably accurately the level of their attainment in English. From the beginning I used a Free Essay [2] type test to make this assessment. I scored these essays on general impressions as most school teachers do. After some time I began to be aware of the dangers of this kind of subjective evaluation - I think I had read Ballard [3] - and began to seek for a more objective method. I evolved something resembling the Performance Inventory which follows a little later on.

More than 30 years later I took up this train of thought again. I had instigated the setting up of a large scale Comparison Trial [4] between two methods [5] of Teaching English as a Foreign Language (T.E.F.L.) in The Gambia in West Africa. One of these methods involved the use of a form of Simplified Spelling. I was, arrogantly enough, not very happy with current modes of assessment which I felt had many weaknesses and so I set about recreating and modifying the objective tests [6] I had used many years before.

I drew up a schedule which I sent off to Professor Laurence Kasdon (Ferkauf Graduate College, New York) for his advice. In the course of our correspondence Prof. Kasdon made several suggestions which were incorporated into the Inventory. Later other modifications were made during consultations with the West African Examinations Council.

The West African Examinations Council had this to say about the Tests in the final report of the trial [7]. "A free-essay test was used. It was believed that for testing English at this rather low level, a 'free essay' test in which children could write on whatever they liked, has important advantages to recommend its use, in preference to the objective (multiple-choice) type test. It is believed that the free-essay reveals the child's real knowledge of the language - he is not unduly stimulated or adversely inhibited or constricted by problematic linguistic materials that the examiner may unwittingly include in a questionnaire or multiple-choice type of test. Besides, it is easy to administer, even a non-literate may supervise [8]. But perhaps more importantly, the usual difficulties associated with the scoring of essay tests are almost completely obviated by the use of a scoring schedule which a relatively unskilled clerical worker can handle since it involves largely very simple, if sometimes tedious, counts and calculations. And although quite highly skilled personnel are needed to make the evaluation, they arrive at the task fresh and unbedevilled by having had to make the initial counts and calculations.

"For the purposes of a comparative evaluation a list of forty-nine variables had originally been proposed, grouped into five categories - essay, kinds of words, sentences, punctuation and general. These were finally pruned down to produce the list (in the Inventory).

"The variables under essay are believed to be a valid measure of verbal fluency; those under kinds of words are taken as indicators of relative sophistication of vocabulary, while those under sentences are measures of overall maturity of language. In general, all the variables are believed to contribute to the assessment of a pupil's proficiency in English as a Second Foreign language."

I as inventor should at this point emphasise that no norms have been established for this kind of test and I am doubtful if a 'norm of universal validity could be established. There are so many factors (age, environment, years at school, etc.) which would have to be considered. It should however be possible to establish local norms for areas such as U.S.A., Great Britain, West Africa and similar territories.

However the inventory, as it is now, is quite suitable for comparing objectively several groups of learners or, indeed, also for establishing objectively the relative position of single learners inside a group.

[1] The learning, teaching or study of English as a Foreign Language.

[2] Free Essay tests are essay tests in which no subject for the essay is set by the examiner and in which the learner may choose his own subject.

[3] Ballard, P.B., The New Examiner. Univ. of London Press, London, 1946.

[4] A comparison trial is one in which usually two or more groups of learners have their scores as groups compared.

[5] Taylor, A., New Nation English. Nelson, London, 1963. Insley, N. and O'Halloran, G., Dynamic English. Pitman, London, 1969.

[6] An objective test is an examination whose results are not influenced by anything except the testee's knowledge.

[7] Osanyimbi, J. et al. Evaluating The Dynamic English Programme. S.S.S. Conference, London, 1975. Photocopies available at £1.00 ($2.50) including postage.

[8] Especially important in a country where distances are great, personnel scanty and schools far between each other.


(Version Agreed with W.A.E.C. [1])

This is a suggested check list or inventory for evaluating the performance of children or groups of children in the use of English. This list is related to a Free Essay type of Timed Test. Only items which may be counted objectively are included. No assessments of the literary merits of the composition are made with this test.

A. Essay.

1. Average length in words of essays.
2. Length in words of longest essay.
*3. Percentage of essays with more than .......... words.
**4. Percentage of essays with less than .......... words.
5. Average number of different words per essay.
6. Greatest number of different words per essay.
*7. Percentage of essays with more than .......... different words.
**8. Percentage of essays with less than .......... different words.

B. Kinds of words.

9. Average number of substantives (nouns, pronouns) per essay.
10. Average number of descriptives (adjectives, adverbs) per essay.
11. Average number of operatives (verbs) per essay.
12. Average number of constructives (prepositions) per essay.
13. Average number of words of one syllable per essay.
14. Average number of words of two syllables per essay.
15. Average number of words of three syllables per essay.
16. Average number of words of 3+ syllables per essay.

C. Sentences.

17. Average number of sentences per essay.
18. Average number of simple sentences per essay.
19. Average number of subordinate clauses per essay.
20. Average number of sentences with 1+ subordinate clauses.
21. Essays without any recognisable sentence.

D. Punctuation.

22. Number of commas correctly used.
23. Number of full stops correctly used.
24. Number of question marks correctly used.
25. Number of capitals correctly used.
26. Number of commas incorrectly used.
27. Number of full stops incorrectly used.
28. Number of question marks incorrectly used.
29. Number of capitals incorrectly used.
30. Number of quotes not used where they should have been.

E. General.

31. Number of mis-spells.
32. Number of syntactic mistakes.
33. Number of groups of words without meaning.

*The number to be inserted in the blank space in Nos. 3 and 7 is that number which is the mean of the number of words of the top 10 essays in each group.

**The number to be inserted in the blank space in Nos. 4 and 6 is that number which is the mean of the number of words in the bottom 10 essays in each group.

[1] W.A.E.C. is the abbreviation for West African Examinations Council which is the statutory examining and research body for the four West African States: Nigeria, Ghana, Sierra Leone and The Gambia.

Back to the top.
On other pages part 1, part 2, part 4, part 5.