1 This research was conducted in a law school teaching clinic because such clinics are excellent sites for empirical research about teaching methods as well as law practice techniques. Five examples of empirical research projects conducted in law school clinics are briefly described in Ronald W. Staudt, Clinical Educators and Empirical Research inAALS Clinical Teachers Conference Handbook at 3–8 to 3–16 (Big Sky, Mont.: Association of American Law Schools, 1980). Reports about three of these clinical research efforts have been published recently: Condlin, Robert J., Socrates' New Clothes: Substituting Persuasion for Learning in Clinical Practice Instruction, 40 Md. L. Rev. 223 (1981); Gary T. Lowenthal, Theoretical Notes on Lawyer Competency and an Overview of the Phoenix Criminal Lawyer Study, 1981 Ariz. St. L.J. __; Sprowl, James A. & Staudt, Ronald W., Computerizing Client Services in a Law School Teaching Clinic: An Experiment in Law Office Automation, 1981 A.B.F. Res. J. 699. Another project received National Science Foundation funding in 1981; a major research study will be conducted in the clinic at the University of Tennessee by Dean H. Rivkin in cooperation with two psychologists. Rivkin's project will examine the problem-solving techniques of lawyers in an attempt to identify the elements of lawyer competence within various areas of legal specialization.Google Scholar

2 Sprowl & Staudt, supra note 1.Google Scholar

3 Lawrence R. Atkin, a computer scientist affiliated with the Northwestern University Vogelback Computer Center and author of several major programs for users of the Northwestern computer, and coauthor Staudt worked together on the design and implementation of this system with assistance from Gary S. Laser, Associate Professor, Illinois Institute of Technology/Chicago Kent School of Law, and coauthor Sprowl. The actual computer programming was done by Atkin. Staudt, with Laser's assistance, identified the information needs of the center and selected the means of meeting those needs which were implemented in the system.Google Scholar

4 Gary S. Laser, Charging Fees in a Law School Clinic: Prepaid Legal Services Programs and In-House Clinical Education, 10 Council on Legal Education for Professional Responsibility, Inc. [newsletter], No. 2, 1977, p. 1.Google Scholar

5 In 1977, 301 new client matters were initiated at the center. In 1978 the number was 1,533, and in 1979, 2,007 new client matters were initiated.Google Scholar

6 The center used a Safeguard card index system to keep track of the status of cases. Multiple copies of cards with client and case data were prepared and separately filed when cases were opened. The center also kept separate files for opening and closing forms. A separate docket system was used to remind lawyers of court dates and deadlines.Google Scholar

7 The City of Chicago cases were referred from five neighborhood offices called Information and Referral (I & R), which are centers operated by the city to provide social services to the elderly population of Chicago. The city was interested in case information that indicated the general type and the number of legal matters accepted from each I & R Center.Google Scholar

8 In addition to these two major sources of clients the center accepted some employment discrimination cases under a contract arrangement with the Equal Employment Opportunity Commission and the Chicago Lawyers Committee for Civil Rights Under Law. In 1979 the center received a grant from the U.S. Department of Health, Education, and Welfare to establish a criminal defense litigation project. The center also accepted clients who needed uncontested guardianships on referral from the Cook County (Illinois) Probate Court. All of these projects required management reports, and several required time analyses.Google Scholar

9 There were very few of these hourly cases when the computer system was being designed. The plans for the development of the center included an increased reliance on individual fee cases. The long-term information needs of the center included the capability to keep track of this information and provide reports for client bills.Google Scholar

10 David Barnhizer, during clinical research study in 1974, required students at the Cleveland State University Law School Clinic to make logs of the time spent on clinic activities. Barnhizer reached the following conclusions about the data-gathering process. As the data were tabulated for each student, clearly identifiable patterns developed. These patterns provided information about the students' basic strengths and weaknesses, particular tendencies in the approach to the legal process, and deficiencies in each particular student experience. The teacher now had access to a detailed source of information about each student for whom he was responsible. This could be used in adjusting and directing the students' clinical learning experience as it was occurring. This knowledge greatly increased the students' learning while, at the same time, the data patterns served as a “check” upon the teacher and as an “early warning device” for students who were experiencing a great deal of difficulty or confusion or were attempting to slide through the clinical course for substantial credit without sufficient input. Barnhizer, David R., The Clinical Method of Instruction: Its Theory and Implementation, 30 J. Legal Educ. 67, 85–92 (1979).Google Scholar

11 These programs are not entirely portable. Assistance from a computer scientist is required to install them on a new computer. The programs are available from Gary S. Laser, IIT/Chicago Kent School of Law, 77 Wacker Drive, Chicago, Illinois 60606. If enough law school clinics are interested, a consortium of clinics could jointly develop an improved, fully portable version of these programs utilizing a uniform set of attorney/student activity codes to facilitate research (see sec. IV infra.).Google Scholar

12 See text infra at pp. 1121–22 for a discussion of the need to develop an efficient manual system before implementing data processing.Google Scholar

13 There are many useful resources to assist lawyers in selecting computer systems to handle administrative tasks in law offices. See, e.g., Mary Ann Altman & Robert I. Weil, How to Manage Your Law Office, pp. 12–1 through 12–39 (New York: Matthew Bender, 1973); Frank Arentowicz, Jr., & Ward Bower, Law Office Automation and Technology (New York: Matthew Bender, 1980); Robert P. Bigelow, ed., Computers and the Law: An Introductory Handbook (3d ed. Chicago: Commerce Clearing House, 1981); Bigelow, Robert P., Where to Go for Help on Law Office Management, 1980 Legal Econ. 25; Bruce D. Heintz & Lavina S. Dill, eds., LOCATE: Law Office Computer Applications Techniques and Equipment (Chicago: ABA Press, 1981); Paul S. Hoffman, Computers for the Law Office—A Survey of Some Available Management/Accounting Financial Systems for the Small-to-Medium-Size Firm, Parts I and II, 21 Law Off. Econ. & Management 72 (1980), 160 (1981); id., ed., Law Office Economics and Management Manual 221–91 (Wilmette, Ill.: Callaghan & Co., 1981).Google Scholar

14 We discuss “on line” data base management systems in our treatment of the selection of data structure, infra at p. 1119.Google Scholar

15 A sample proposal will illustrate this point. One well-known computer bureau offered to program a data base system suited for our needs with the exception that no data would be “on line” except for one short period at the end of the month. The bureau estimated that such a system would cost approximately $1,500-$2,000 each month. This was a staggering amount of money in the face of our grant and operational budgets. The system that was ultimately installed costs less than 10 percent of this monthly charge. The main advantage of the commercial systems was the speed with which they could be implemented. Either commercial alternative would have been operational much sooner than a system written expressly for our needs.Google Scholar

16 MELS is a union prepaid legal services plan that had developed a computerized case and time management system to report on the activity of its staff and clients. The volume and type of clients at MELS were similar to those of the IIT/Chicago Kent College of Law Legal Services Center, and the MELS system could have been installed on the University of Chicago's IBM computer, which our terminals could contact by telephone.Google Scholar

17 The MELS system condensed its data in a way that did not preserve the individual detail records, and the condensed data could not be readily extracted from storage for processing by standard statistical programs. The MELS system is very efficient for storing large masses of data but difficult for another programmer to decipher. Individual detail records are broken up and stored in client-organized storage slots. This organization makes it hard for a standard statistical package to unravel the data and rearrange them in some new order for research purposes.Google Scholar

18 See note 3 supra for explanation of Atkin's background and his role in the project.Google Scholar

19 Sprowl & Staudt, supra note 2, at 721–23. The Control Data Corporation 6600 computer that we used to design and implement the system described here is a large scientific computer located at the Vogelback Computer Center of Northwestern University in Evanston, Illinois. Other experiments with computer technology under way at the center required that we be able to communicate with this particular computer, and our staff was required to learn how to use many of the “utility” programs available at the Vogelback Computer Center. Id. at 714–15. Even though the work done by this computer for our management reports could have been accomplished on a much smaller computer, there are major advantages in using a large computer at a university computer center, including the reduced cost of computer time for educational users and the availability of expert consultants who are interested in experimental use of computers. There are also disadvantages in using university computer centers. Id. at 715–16.Google Scholar

20 See CODASYL Systems Committee, Selection and Acquisition of Data Base Management Systems (New York: Association for Computing Machinery, March 1976). For a detailed description of a generalized data base management system and the data structures it supports, see Data Base Task Group Report to the CODASYL Programming Language Committee (New York: Association for Computing Machinery, October 1969). CODASYL is an industry and government group that developed and now maintains and supports the business-oriented programming language COBOL. The above references relate to CODASYL's efforts to tie generalized data base management systems directly into the COBOL language.Google Scholar

21 See text at note 14.Google Scholar

22 This “column” and “line” convention grew out of the traditional use of punched computer cards to store data. With such cards, specific types of data are stored in specific “columns” of each card, and each card represents a single transaction. Much financial information is still stored in this fashion, although the use of punched cards is gradually being replaced by the use of the “buffered” terminals of the type described in this article. For an overview discussion of the use of sequential data files, see Thomas R. Gildersleeve, Design of Sequential File Systems (New York: Wiley-Inter-science, 1971).Google Scholar

23 The center uses four magnetic tapes to store data so that back-up tapes are always available. Each tape contains data collected in previous years and data summaries produced in report processing. The tapes cost $2.00 per month to store, with no extra charge for length. Northwestern University's computing center staff mounts and unmounts these tapes in response to commands issued automatically by the management system. The Legal Services Center staff had no need to see any of these tapes. Most university computing centers provide comparable tape-mounting services.Google Scholar

24 See text supra at notes 16 and 17 describing our consideration of use of the MELS system. The MELS reports are described in Helen D. Johnson, MELS Systems' Procedures (2d ed. New York: District Council 37 Municipal Employees Legal Services Trust, 1978). A copy of the manual is on file at the IIT/Chicago Kent College of Law Library.Google Scholar

25 During the time of this experiment, the Legal Services Corporation was testing a computer-driven case and time management system at selected poverty law programs across the country. The first version of this system, the Project Reporting System (PRS), met with great resistance in the poverty law offices. The second attempt, called the Statistical Reporting System (SRS), was less offensive to poverty lawyers, and the system was tested in selected poverty programs across the country. Both systems were written and implemented by Abt Associates, Inc., Cambridge, Massachusetts.Google Scholar

26 See appendix B for a description of how a line of data in the case record is organized.Google Scholar

27 See appendix C for a description of how a line of data in the time record is organized.Google Scholar

28 The most difficult task in writing this system was selecting the activity codes. The codes had to be easy to use yet comprehensive enough to gather detailed information. These two goals were clearly in conflict. Using a large number of codes helps to classify activities in detail, but it also inhibits easy use because the large number of codes are difficult for lawyers and students to remember. A short code list is simple for staff members to remember and use, but it does not record much detail about the activity performed. A small number of codes may prove to be more accurate than a large number, however, because the small number of codes are more likely to be used properly. A moderately complex activity code with 37 separate categories was ultimately adopted as a compromise. See appendix D for a description of these codes.Google Scholar

29 These terminals were Teleray 1061 microprocessor-controlled CRTS, costing just over $1,000. The terminals provided buffered storage of data and permitted “off line” editing of data entered onto the screen, using terminal editing controls. Through the use of protected field delimiters, these terminals could provide the user with questionnaire forms in which the forms were “protected” from user editing while the blanks in the forms could be filled in and later revised by the user. A “block transmit” key allowed the unprotected data entered by the user to be transmitted to the computer as a unit for central processing and storage. Special function keys on the terminal could be “programmed” by the central computer to transmit special messages back to the central computer whenever these keys were used. Our system made use of all these features of the Teleray terminals. Comparable terminals can be obtained from many different suppliers, and terminals such as these can greatly simplify the process of designing an automated data-gathering system. An excellent terminal that can display a questionnaire form, accept data entries, and then transmit only the data entries back to a central computer is the VT-132 terminal manufactured by Digital Equipment Corporation. The VT-132 has a 12-inch display. A terminal that is compatible with the VT-132 but that offers a 15-inch display and user-programmable special function keys is the Teleray Model 100. Hazel-tine's Executive 80 terminals also include provision for data gathering and provide a 15-inch display and special function keys. All the above terminals are at least partially compatible with American National Standard ANSI X3.64-1979, which sets forth international standard protocols for computer-to-terminal communication. These terminals may therefore be used for other purposes, and they are unlikely to become obsolete. See, generally, the technical manuals for these terminals to obtain an explanation of how these “block mode” or “buffered” or “data entry plus editing” terminals operate. The Teleray Model 1061 terminals that we used are not presently ANSI X3.64-1979 compatible.Google Scholar

30 Casein and Timein compare each new data line with another record called “INFORM,” which contains lists of the permitted abbreviations for each data category. INFORM contains the name and initials and some demographic data about each lawyer and each student who works in the center. Casein compares the attorney initials in columns 26–28 of a case record with the list of “legal” initials contained in the INFORM file. Any combination of letters or numbers or blanks in columns 26–28 which does not match a set of legal initials is flagged as an error. Errors that relate to prior data records cannot be detected at this time. For example, a case number could be transposed to mistakenly attempt to close a closed case or to enter time to a nonexistent case. These errors cannot be detected until the historical data file is attached from tape storage at the end of the month. A data base management system would facilitate both kinds of error checking at the time of data entry because the historical records are always available in active memory for comparison.Google Scholar

31 A sample of the sequence of prompts and replies that occurs when a staff member runs the system to obtain monthly reports is set forth in appendix A.Google Scholar

32 Data errors appear in an “exception report,” which is printed automatically after the computer is furnished with properly labeled input files. The exception report is compiled by comparing data taken from the historical files with the current input files and with lists of permitted changes in personnel. Such things as time entries for nonexistent cases are flagged as errors. Once the exception report is printed, the report-generating job is usually halted so that the necessary corrections can be keyed into the current input files before the monthly reports are written. Several lines from a typical exception report are set out in appendix E.Google Scholar

33 If all the reports are to be typed, the printing process on a typewriter terminal operating at 30 characters per second (cps) can require 5 to 6 hours; more than 300 pages of reports are written. Usually only eight or nine of the reports are selected, and the process is correspondingly quicker. The forms tractor permits the operator to do other work in the same room without attending to the terminal to change pages. The “daisy wheel” printer (Anderson Jacobson Model 832) operates at about 30 cps. Also available are “Dot Matrix” printers that work at 120 cps, which would cut the time required to generate these reports to one-fourth. Had Northwestern's computation center been located near the clinic, we could have had these reports printed by the center's high speed printers. Many university centers now use a laser printer to print lengthy reports on 8½× 11-inch paper at extremely high speeds.Google Scholar

34 When the computer prints Detail Reports 4 and 5, the most recent case record for each pending case is printed in full. The programming for these reports is very simple because the computer merely identifies the proper lines in the case record, sorts the lines into alphabetical order, and then prints the lines sequentially. The first few lines of a typical Detail Report 4 are printed below:Google Scholar

All the reports illustrated in this article are altered to omit or mask client names to protect against any breach of confidentiality.Google Scholar

35 The Dormant Cases report is more sophisticated than the alphabetical lists. This report combines case data and time data in one report. The computer compares pending case records with the time records to find those case records for which no time was entered during the relevant period. The report prints the case data on the left side and the data calculated from the time records on the right side.Google Scholar

36 Detail Reports 1, 2, and 3 each contain the same lines of case and time data, but the report lines are sorted differently in each report. Detail Report 1 is written for the individual attorneys; the cases are sorted alphabetically for easy reference by the lawyers. In Detail Report 2, cases are sorted by attorney and then by type of legal problem so that clinic administrators can monitor case concentrations and help match case assignments to attorneys' specialties and experience. Detail Report 3 sorts the same data lines into the projects in which the case originated, and then cases are arranged within each project by the type of legal problem presented. This report is used by project directors to manage and monitor resources devoted to each project. A sample of Detail Report 1 is set out in appendix F.Google Scholar

37 The center's prepaid plan allowed each covered employee 15 hours of attorney time without any fee. Hourly fees of $15 per hour were imposed for the next 10 hours; regular hourly rates were charged for each additional hour. The city of Chicago contract also contained coverage limits for attorney time. These limits on time for individual cases were not monitored carefully until the computer system was installed; manual accounting for thousands of clients was too expensive. Detail Report 3 provides this information organized conveniently for the project directors of the prepaid plan and the city of Chicago project.Google Scholar

38 Many client matters handled by the center for fees were small cases for moderate-income clients which required only two or three hours of attorney time. Detail Report 1 tallied the amount of time spent and listed the type of case or legal matter. Hence, a simple bill could be prepared from this report.Google Scholar

39 These reports were designed by Staudt and Atkin to match manual case accounting reports previously made for the center's director. Although most of the information in these tabular summaries can be obtained by reviewing Detail Reports 1, 2, and 3, the center's director was accustomed to this summary report format. The computer system provides these reports structured to match prior manual tallies so that case concentrations can be monitored without a change in reporting style. These reports are good examples of the flexibility available for a law office if it develops its own programs. Within cost limits, data can be reported in several different styles to meet the individual preferences of law office managers.Google Scholar

40 The telephone screening code had been changed several times to improve its ability to provide relevant information about the value of this service to the senior citizen clients. One reason for providing the service is to help avoid unnecessary trips by senior citizens to distant neighborhood centers. Therefore it is important to monitor the effectiveness of the telephone service in resolving client problems.Google Scholar

41 Closing codes have been in use since the early days of the federally funded legal services program under the Office of Economic Opportunity (OEO). Poverty law offices funded by OEO were required to submit periodic reports on case results, and a code was used to identify the type of legal services performed for the client. When detailed time data are available, the closing code information can be less valuable. Our closing code report and the telephone screening report were designed with numerical code lists so that the data categories could be changed without any reprogramming. The most recent descriptions assigned to the closing code numbers are set forth in appendix I. Revision of this list is already considered necessary by the center lawyers because the categories do not provide enough useful information about the criminal defense cases that are becoming a large part of the work of the center.Google Scholar

42 The Activity Profile is explained in some detail in the text. Summary Report 6, the Project Profile, is derived by the computer from the same time records. The format of time records is set out in appendix C. Summary Report 6 is also divided into parts, each of which sets out time and activity information in a cross-hatch format. The city of Chicago Senior Citizens contract is the largest source of the center's clients, and the first part of the Project Profile reports on senior citizen cases to provide information about attorney and student resources expended for each of the city of Chicago neighborhood intake offices (I & R Centers). The report also identifies time expended for general administration for the city not related to a specific client. The second part of the Project Profile reports the same type of information for each project at the center. An example of the second part of the Project Profile is set out in appendix K. The report is used to monitor the amount of resources expended on each project for grant accountability and resource management within the center.Google Scholar

43 The attorney activity information reported here is identical to the data reported in the Attorney Activity Profile. In the Attorney/Student Activity Profile each attorney is reported on a separate page together with time summaries for each of the attorney's students.Google Scholar

44 The INFORM file for fall 1979 contains a list of all of Strickler's students together with some data about each student, such as the number of semesters of law school completed, the number of clinic courses previously completed, and so forth. A unique identification code is assigned to each student. If possible each student's own initials are used as part of the identification code, preceded by a letter drawn from the initials of the supervising attorney. Therefore, in table 1, column 2, “RTS” could have been derived as follows: R = identification of supervising attorney + TS = the initials of student “Tony Studious.”.Google Scholar

45 This extension of the system is further illustration of the value of the command module described in appendix A. Another computer programmer, Steven Keyser, wrote the programs to gather the financial data, process the data, and prepare accounting reports and client bills. Keyser used the command module to integrate this financial accounting system with our other computer systems so that data entry, correction, and revision required very little staff training. See Sprowl & Staudt, supra note 2, at 721–23.Google Scholar

46 Boyer, Barry B. & Cramton, Roger C., American Legal Education: An Agenda for Research and Reform, 59 Cornell L. Rev. 221, 279, reprinted in Research Contributions of the American Bar Foundation, 1974, No. 1. One result of this recommendation has been the ABF Law Student Activity Project and Ronald Pipkin's series of analyses using time and activity data he collected at 14 different law schools. See Ronald M. Pipkin, Law Schools and Professional Education: Report and Recommendations of the Special Committee for a Study of Legal Education of the American Bar Association (Chicago: ABA Press, 1980). Pipkin's data were daily diaries prepared by 1,200 students representing three class years at seven law schools. Id. at 39. Pipkin spent years classifying and sorting these data into categories to draw meaningful conclusions. Several articles have been published about these conclusions, and a book is forthcoming. See id., Legal Education: The Consumer's Perspective, 1976 A.B.F. Res. J. 1161;id., Student Responses to Law School Curricula (paper presented to the Section on Legal Education and Admissions to the Bar, at the Annual Meeting of the American Bar Association, Chicago, August 1977); id., Law School Instruction in Professional Responsibility: A Curricular Paradox, 1979 A.B.F. Res. J. 247;id., Moonlighting in Law School: Attitudinal and Behavioral Correlates of Part-Time Employment by Full-Time Students, A.B.F. Res. J. (forthcoming); id., Law Student Activity Patterns (book in preparation for publication by the American Bar Foundation, 1982). Pipkin's study of student activity patterns is a long first step along the path of inquiry suggested by Boyer and Cramton. In 1974 and 1975 David Barnhizer also collected some time data in the form of daily logs from 62 students enrolled in clinical law courses at Cleveland State University. The daily logs charted 14 categories of clinical activities. Barnhizer compared the student activity data with the number of credit hours awarded for the clinical course and predicted that clinical courses would be most effective if more than 5 credits and less than 12 credits were awarded in one semester. Barnhizer also studied time demands and task frequency demands on clinical teachers. Barnhizer, supra note 10, at 83–92.Google Scholar

47 Boyer & Cramton, supra note 46, at 281.Google Scholar

48 Id.Google Scholar

49 Some costs associated with using this system to maintain a data base of student time are properly ascribed to the research objectives of a clinical program. It is expensive to key in all the data even using the data entry program designed to facilitate this process. The center has not entered student time data for every student in every semester. There are feasible methods of reducing this data entry cost if a sufficient volume of information is to be entered. Optical scanning readers are one such solution for reducing data entry costs for large masses of information. As law practice moves toward a paperless office and law schools are forced to teach law office technology, students may be sophisticated enough with computers to enter in their own time.Google Scholar

50 Our coding system was also arranged to identify the total amount of time that is spent in computerized client services and in computer system design. For example, the computer code categories enabled us to estimate how much time Robert Seibel spent building the center's experimental willdrafting system. See Sprowl & Staudt, supra note 2, at 727–29. See appendix D for an explanation of each code category.Google Scholar

51 See, e.g., Strong, Frank R., The Pedagogic Training of a Law Faculty, 25 J. Legal Educ. 226, 230–31 (1973). Another possible source of relevant activity descriptions is found in Frances Kahn Zemans & Victor G. Rosenblum, The Making of a Public Profession 124–25 (Chicago: American Bar Foundation, 1981).Google Scholar