WO1982001613A1

WO1982001613A1 - Diagnostic and remedial method for learning disabilities

Info

Publication number: WO1982001613A1
Application number: PCT/US1981/001486
Authority: WO
Inventors: Marion C Bossart
Original assignee: Marion C Bossart
Priority date: 1980-11-05
Filing date: 1981-11-03
Publication date: 1982-05-13
Also published as: EP0064549A1; BE891829R; BE891010A; EP0064549A4; OA09582A

Abstract

A method for diagnosing learning disabilities such as dyslexia, and other visual, auditory or perceptual defects which are manifested by misspelling and vocabulary deficiencies; and means for palliating such disabilities comprising the processing of a text (TXT) written by an affected subject through a programmable machine (12) which compares each word to a basic vocabulary list (BWL), classifies it as correct or misspelled, and accumulates statistical data about the subject's deficiencies. The programmable machine is also made to generate a spelling dictionary (DIC) which is used to correct the common spelling mistakes of the subject. A text written by the subject which otherwise would be barely legible can be quickly upgraded; thus expanding the subject's capability to communicate in writing.

Description

Description Diagnostic and Remedial Method far Learning Disabilities Prior Application

This application is a continuation-in part of United States Application No. 203,552 filed 11/05/80. Background of the Invention Individuals with dyslexia experience serious problems processing language although they are of average or above average intelligence. Their impairment is of unknown etiology, and usually manifests itself as a dysfunction in the visual and/or auditory perceptual-memory proces. A conspicuous sign of their disability is their poor memcry of the graphic configurations of many words. Consequently, these individuals wϋl tend to use a phonetic spelling with a certain degree of consistency throughout their writings. Their works are often barely legible to someone who is not familiar with their unorthodox spelling, It would be unfortunate if some of those individuals could not share their ideas in writing with others, except in a specialized scholastic environment.

Such spelling disabilities also tend to severely affect the choice and use of words. Consequently, a resected word use may often be observed in such individuals in contret with a normal or above average level of syntactic maturity. Accordingly, the careful analysis of a text written by a poor speller may lead to the diagnosis of dyslexia or other types of leaning impairment.

Furthermore, the analysis of a series of such texts can provide the material necessary to establish a personalised spelling dictionary which can be made available to the subject to automatically, or on demand correct his spelling mistakes. A text which otherwise would be barely legible can be quickly upgraded; thus expanding the subject's ability to com municate in writing. The analysis of such an orthographically deficient text can be accomplished most efficiently using program mab le machines which can provide prerecorded answers in response to certain stimuli or combinations thereof. The present invention can make use of either a standard processor com mercially available under various trademarks or of a specialized program mable machine dedicated to these tasks.

Summary of the Invention

One of the principal objects of the invention is to provide a method by which a text written by a subject with learning deficiencies can be analyzed in order to derive statistical data on the subject's vocabulary and orthograph helpful in the diagnosis of dyslexia and other learning disabilities.

A further object of the invention is to provide a method for documenting a subject's spelling idiosyncrasies and to create a spelling dictionary which can be used in the correction of subsequent texts written by the same subject.

Another object of the invention is to provide a means by which written com munications from individuals affected with auditory and/or visual perceptual disorders can be quickly and automatically upgraded into an easily legible form; thus expanding the individual's ability to com municate.

These and other objects are achieved by a method in which a text written by a subject is mechanically analyzed in order to classify the words of the text by reference to basic vocabulary lists and a misspelled word dictionary which are periodically updated in function of new text entires. The misspelled word dictionary is also available to automatically provide the correct spelling of words in response to the idiosyncratic version presented by the subject. Statistical data on the number and type of words found in the text are continuously accumulated in order to provide a vocabulary profile of the subject which may be useful in the diagnosis of certain learning disabilities. Brief Description of the Drawings

Figures 1A through 1E show the flow diagram of the diagnosis process wherein:

Figure 1A illstrates the analyzing process applied to the raw text entered by the student; Figure IB illstrates the computation of the diagnosis indice ; Figure 1C illstrates the analysis process for the intermediate file;

Figure ID illustrates the handling of new words;

Figure IE illustrates the process of automatic misspelling error correction;

Figure 2 is a general block diagram of the apparatus used in the claimed method;

Figure 3 is the general block diagram of an alternate embodiment of the apparatus; Figure 4 is a diagram of the data flow between the various files;

Figures 5A through 5H show the structure of the program used to control the apparatus wherein:

Figure 5A illustrates the top level of the operator selection process;

Figure 5B illustrates the text entry process;

Figure 5C illustrates the text printout process;

Figure 5D illustrates the printout of the files;

Figure 5E illstrates the printout of the diagnosis indicia; Figure 5F illustrates the selection of the corrective proceeses; Figure 5G illustrates the inter mediate-to-final text process;

Figure 5 H illustrates the calculation of the indicia;

Figure 6 is a modified block diagram of the first embodiment illustrated in Figure 2; Figure 7 is an alternate version of toe program selection shown in Figure 5C illustrating the output of the final text in specialized form; and

Figure 8 illustrates the printout of oversized characters by combination of symbols. Descripti on of the Preferred Embodiments of the Invention

Referring now to the drawings, a diagnostic method shall first be described for establishing two indicia of lexical proficiency by analyzing a particular text written by a student with learning disabilities. The first indicium is the word TYPE to TO KEN RATIO

(TTR), which is the ratio of the number of original, words found in the text after discarding all repetitions; over the total number

Wf words including any repetitive use of the same word.

The second indicium is the BASIC ER R O R INDEX (BEI), which is the ratio of misspelled words belonging to a predetermined BASIC WO RD LIST (BWL); over the total number of misspelled words found throughout the text.

The TTR gives a general indication of the student's vocabulary diversity. The BEI reflects the student's inability to write accurately even com mon words.

The flow diagrams of the figures 1A and 1B illustrate the various diagnostic steps. In the flow diagram circles represent files in which a list of alpha numerical words can be stored. Diamonds represent decision-making steps. Squares represent numerical counting functions and rectangles denote various other operations. The process begins with the entry of a basic word list into the BWL file. This basic word list is representative of the standard vocabulary which a student within a particular age bracket is expected to know and use commonly.

The next step is the entry into a dictionary file DIC of the common spelling mistakes known to recur in the particular student's writings, along with the correct spelling cf the words.

A third step, optional at that time, is to enter into a personalized word list file PWL some basic vσrds which while they are not found in the basic word list yet are well-known and commonly used by the particular student. Next, the student's raw text is entered into the student's raw text file TEX. The analysis of the text file is done word per word. The total number of words are counted and the numerical count is kept in the token register TOK. The word is first compared to the words in the B WL file. If the word is found to be in the BWL file it is transferred im mediately to the intermediate text file INTER. If the word is not found in the B WL file, it is next compared to the words in the P W L file. If the word is identified as one which appears in the P WL file, it is transferred to the INTE R file. If the word is not found either in the BWL file or in the P WL file, it is then checked against the misspelled words contained in the DIC file. If the word cannot be found in the misspelled word list of the DIC file, it is counted as an unknown word and the numerical count is kept in the unknown U NK register. That word is also entered in the NFD file and into the INTER file. If the word is recognized as one of the co mm only misspelled words listed in the DIC file, then the corresponding proper spelling is fetched from the DIC file and subsequently entered into the INTER file. The number of identified misspelled words is accumulated and the numerical count is kept in the MIS register. The correction fetched from the DIC file is also compared to the words in the BWL file. If correspondence is found the number of misspelled wards from the basic wcrd list is incremented by 1 and the numerical accumulation is kept in the MBW register. If the corrected spelling is not of a word found in the BWL file it is used to update the personalized wcrd list by entering it in the PWL file.

As words are entered into the INTER file, they are compared to all the words that have previously been entered therein, and are counted if they are encountered for the first time. The numerical count of newly encountered wcrd is kept in the TYP register.

At this point in the process the student's raw text has been updated by replacing all the misspelled wcrds which have been identified with their correct configurations derived from the misspelled word dictionary file DIC. All the unidentified words have been stored in the NFD file, and the PWL file has been updated by adding into it words which are not found in the basic word list but have been used by the student. The TO K register holds the total numbers of words used in the text. The MIS register holds the total number known to be misspelled. The MBW register holds the total number of basic word list words which have been misspelled by the student; and The UNK register reflects the number of unknown words which have been found throughout the text,

As illustrated in Figure IB, a partial type to token ratio TTR is obtained by dividing the contents of the TIP register by the difference between the contents of the TO K register and the UNK register.

A partial basic error index BEI is obtained by dividing the contents of the MBW register by the contents of the MIS register.

This second phase of the process completes the diagnostics. In a third phase of the process, both the misspelled dictionary and the persona lized word list are updated by analyzing the unidentified words contained in the NFD file. This third phase is illustrated in Figure 1C. If the word extracted from the NFD file is found to be correctly spelled, it is im mediately added to the P WL file. If by contrast the word is found to be misspelled, then both the misspelled version and the correction are stored in the

DIC file.

The text contained in the INTE R file can now be updated through a fourth and final phase of the process which is illustrated in Figure ID.

Each word read out of the INTE R file is first compared to the words of the basic word list held in the BWL file. If correspondence is found the word is im mediately entered into the FINAL file. If the word is not found in the B WL file, it is next compared to the words in the P WL file. If correspondence is established, the word is transferred to the FINAL file. I f on the contrary the word is not found to be one stored in the P WL file it is compared to all of the misspelling contained in the DIC file. If the word is identified as a com mon misspelling the correct spelling is fetched from the DIC file and entered into the FINAL file. Words that are not found in either the BWL, PWL or DIC file are discarded as erroneous.

At this point the FINAL file contains a corrected version of the text initially written by the student.

The diagnostic and corrective method just described is better practiced with the use of a programmable machine which can be made to automatically and repetitively perform various required steps for each word of the student's raw text entered therein. As illustrated in Figure 2, such programmable machine must include at least an input device such as a keyboard 3, a programmable processor 2 and an output device such as a printer 3. The keyboard and printer may be combined in an automatic, remotely controlled typewriter or the complete process can be modified and adapted to a similar unit containing a microprocessor. This latter unit would probably serve only as a student's station, however.

The program mable processor 2 may be constituted by a standard data processor with a basic architecture comprising a program memory 4, a control unit 5, operating register 6 and a file memory 7. The control unit 2 operates in accordance with directives received from the program memory 4 to manipulate data stored in file memory 7 through the operating registers 6. Such programmable machines are well-known to these skilled in the electronic arts and need not be described in mere detail.

The repetitive use of such a machine in applying the diagnostic and corrective method to the same individual will automatically result in the accumulation in the file memory 7 of a complete list of words com monly used by the subject in either their correct form (BWL and P WL files); or their particular configuration favored by the student's (DIC file). The machine can then be used in the application of a remedial or corrective method for helping the student in the proper spelling of words. This method is illustrated in the flow diagram of Figure 1E. The method generates one of the three following indications either in printed form or other type of visual display. a. The word is correctly spelled; b. The word is incorrectly spelled accompanied by a display of the correct spelling; and c. The word is unknown.

Each word as it is entered into the machine is compared first to the words in the BWL file and to the P WL file. A positive identification is translated as a correct message indication. If the word is not found in one of those two files, it is next compared to the misspelled words in the DIC file. If the word is not found in the DIC file, an unknown indication is given. If the word is recognized as one of the com monly misspelled ones a misspell indication is given to the student and the corresponding correct spelling is fetched out of the DIC file and displayed to the student.

A partial configuration of the programmable machine designed to practice all the methods described above, and which is particularly adapted to a schooiroom environment is illustrated in Figure 3. The processor 2 which can be locally or remotely installed, has associated with it an automatic typewriter 8 which serves as the teacher's staticn. The processor 2 services a plurality of student stations 9-9 N, each comprising a keyboard 11- 11N and an output device 10-10 N which is preferably an alpha-numerical and/or specialized alphabet printout. The automatic typewriter 8 is used in establishing the various files and for entering the raw student text. One of each file type (except for BWL file which may be shared by several students of the same age group) must be dedicated to each student. Once sufficient data has been accumulated in the various files for each student in the group, the student's station can be used to verify or obtain the correct spelling of the word according to the process illustrated in the flow diagram of Figure 1E.

In the preferred embodiment of the invention the program mable machine may be implemented with a standard data processing machine such as the Micronova model manufactured by Data General Corporation. This machine can be easily program med in the BASIC or any other standard program ming language according to principles well-known to those skilled in the art by reference to the flow diagrams of Figures 1A through 1H and the additional data provided hereinafter.

The interactive functions of the files during the various phases of the process above-described are illustrated in the diagram of Figure 4. The four basic files shown on the left side of the diagram: TEX, BWL, DIC, and P WL are first established then used automatically by the processor to create the INTE R and NFD files. In the intermediate phase of the operation under control of the teacher the DIC and PWL files are updated to reflect the newly discovered words in the student's raw text. In the last corrective phase of the process the text held in the INTE R file is corrected by reference to the DIC , P WL, and B WL files in crder to obtain the final corrected text.

The various tasks of the machines are presented to the operator in a menu-type fcr mat which facilitates their orderly selection. The various functions and tasks are graced into three separate levels. Each level is presented to the operator in the form of multiple choice in successive steps until a specific lower level function is selected. The selection structure is illustrate in the diagrams of Figures 5 A through 5H.

At the top level of the menu structure the operator is asked to select one of the following tasks: 1 Enter student text; 2. Print text (TXT, INTER, and FINAL) files;

3. Print reference (BWL, PWL, DIG and NFD) files;

4. File analysis;

5. Reference file maintenance;

6. Inter to final file update; 7. TTR and BEI index calculation.

The selection of one of the seven top level tasks leads to the corresponding second level menu structure inquiry and so on until the specific task is identified. It should be noted that all the functions illustrated in the flow diagram of Figure 1A are accomplished under the text entry routine shown in Figure 5B, including the calculation of the total were count TOK, the misspelled count MIS, the misspelled basic word count MB W, and the total number of unknown words encountered UNK. These counts can be called out and printed according to the routine shown in Figure 5E. Any file can be printed en request either under the routine illustrated in Figure 5C for the text files, ar according to the routine illustrated in Figure 5D for the reference files. The reference file maintenance routine illustrated in Figure 5F can be used by the teacher when updating the P WL and the DIC fϊLes after his own analysis of the unknown words printed out from the NFD file in order to achieve the functions illustrated in Figure IC. Calculation of the final TTR and BEI indexes may be done on the basis of the corrected text held in the final file by re-entering the contents of FINAL into the TXT file and reprocessing it according to the first, second and third diagnostic phases described earlier. It should also be understood that once extensive files have been accumulated fcr a particular student after numerous text entry and correction passes these files could be part of a small read only memories (RO M) which could be incorporated in a portable unit such as a typewriter or one similar to those illustrated in Figure 3. but which could operate independently from a large scale processor.

The system can be adapted for students with serious visual handicaps by converting the final text to Braille or another specialized alphabet. On the input side it is sufficient fcr this purpose to impress the key of the keyboard Il with kineasthetic symbols. A reassignment of key locaticcs may sometimes be indicated in order to accom modate a particular type of disability. In order to generate a copy of the final text in a specialized alphabet, the output routine illustrated in Figure 5C can be modified as shown in Figure 7 or the system can be equipped with a special alphabet device 13 in addition to a standard alphabet device 14 as shown in Figure 6 or other output device which can generate a copy of the text in specialized form.

In this modified system the data processing takes place in a standard character form. The Intermediate Text is also outputad in standard form. The translation into the special alphabet version takes place in the last stage of output, i e., within the specialdevice 13 or just before output, on a standard device in accordance with the flow diagram of Figure 7. The final text can be printed in both standard and specialized character on alternate lines.

This process could be followed in order to generate geometric characters or extra, large characters using a combination of standard types as illustrated in Figure 8.

The student would have the option to output the final text in special alphabet style, in standard form or an interlined combination of both.

While I have described the preferred embodiments of my invention and suggested variatiors thereof, other embodiments could be designed to apply my method without departing from the spirit of the invention and the scope of the appended claims.

Claims

Claims A method for diagnosing and palliating the spelling deficiencies of a student comprising the steps of: a. entering into a first file a first text written by the student; b. entering into a second file words known to be com monly misspelled by the student in both their misspelled versions and their correct spelling forms; c. entering into a third file a basic vocabulary list of words; d. comparing each wcrd of the text to the words in the second and third files; e. consequently to said com paring, and for each word in the text found to be similar to a misspelled word version in the second file, update the text by substutiting the corresponding correct spelling form fetched from the second file; f. counting the total number of word types, excluding any recurrence of the same word, found in said updated text; and g. computing the ratio of the total number of word types over the total number of words in the text found to be similar either to a word in the second file or to a word in the third file.

2. The method claimed in Claim 1 which further comprises the steps of: h. checking each said correct spelling fern fetched from the second file against each word of the third file; and i. consequently to said checking, calculate the ratio of the number of said correct spelling forms found in the third file over the total number of said correct spelling forms fetched from the second file.

3. The method claimed in Claim 1 or in Claim 2 which further comprises the steps of: j. analyzing the unknown words found in said updated text which are not found to be listed in the second or third files; k. consequently to said analyzing, entering into the three-N file unknown words found to be correctly spelled; l. adding to the second file these of said unknown words found to be misspelled in both their misspelled versions and their correct spelling forms; and repeating steps d and e.

4. The method claimed in Claim 3 which further comprises the steps of: n. entering the once updated text under step e into a fifth file; and o. entering in the twice updated text under step m into a sixth file.

5. The method claimed in Claim 4 w herein step d comprises the step of: p. entering into a seventh file the words from the text found not to be corresponding to words in either the second or the third file.

6. The method claimed in Claim 5 which comprise s steps of repetitively and sequentially: q. clearing the contents of the first, fifth, sixth and seventh files; r. entering a new student text into the first file; and s. repeating steps c, e, j, k, l, m, n, and c.

7. The method claimed in Claim 3 which further comprises: repeating steps f, g, h and i.

8. The method of Claim 3 or 6 which further comprises: converting each standard character in the text to a new pattern formed by the combination of symbols.

9. The method of Claim 3 or 6 which comprises: converting standard characters in the text to special symbols.

10. The method of Claim 3 or 6 which further comprises: converting the updated text into a idneasthetic form perceptible by a student with visual disability. Il. The method of Claim 5 or 5 which further comprises: t. in a preliminary phase, preparing a series of programs designed to cause a data processor to automatically perform said steps on com mand; u. for each application of the method com manding said data processor to execute said programs.

12. The method of Claim H wherein said data processor comprises a microprocessor.

13. The method claimed in Claim 5 which comprises the steps of repetitively and sequentially: v. clearing tee contents of the first, fifth, sixth and seventh files; w. entering a new student text into the first file; and x. repeating steps d, e, j, n and o.

14. The method of Claim 3 or 13 which further comprises: entering into an eighth file a special alphabet; converting each standard character in the text to a new pattern formed by the combination of symbols corresponding to the special alphabet.

15. The method of Claim 14 which further comprises converting each standard character to a Kidneasthetic form perceptible by a student with visual disability. 16. The method claimed in Claim 13 which farther comprises converting standard characters in the text to special symbols.