GSLIS Career Services Office Job-Postings Process and Design: Analysis and Recommendations

Transcription

1 GSLIS Career Services Office Job-Postings Process and Design: Analysis and Recommendations Prepared by Students in the Summer Session 2000 Class in Systems Analysis and Evaluation Jie Chen Amy Gannon Yu-Lu Hsiung Jia Le Steven Schang Judi Tidwell Stacey Wuest Under the direction of Professor Ronald E. Wyllys Graduate School of Library and Information Science The University of Texas at Austin 2000 July 12

2 Those copyrighted items (if any) reproduced herein have been copied as fair use for the educational purpose for which the study treated herein was undertaken and is here reported. Except for such otherwise copyright items, the entire contents of this report are copyright 2000 by Ronald Eugene Wyllys for the benefit of the GSLIS Foundation, Graduate School of Library and Information Science, The University of Texas at Austin, Austin, Texas Permission to copy, quote, and/or reproduce, and/or use all or any part of the contents of this report (except those parts that are otherwise copyright) is granted to all persons and organizations, provided that due acknowledgement is made of the source of the material thus utilized. Telephone: /2742; Fax: Graduate School of Library and Information Science Sanchez Building 564 The University of Texas at Austin Austin, Texas ii

3 Table of Contents 1. EXECUTIVE SUMMARY SUMMARY OF IDENTIFIED REQUIREMENTS SUMMARY OF RECOMMENDATIONS INTRODUCTION MISSION STATEMENT OVERVIEW ANALYSIS SYSTEM HISTORICAL DATA CURRENT CSO OPERATIONS AND ORGANIZATION SYSTEM ENVIRONMENT FOR JOB POSTINGS SURVEYS Survey Results Student Interviews Faculty Interviews Career Services Interviews Desired Capabilities RECOMMENDATIONS DISCUS AS A JOB-POSTINGS SOLUTION Discus and Identified System Requirements Limitations of Discus Alternative Solutions PROPOSED DISCUS DESIGN Job Categories and Title Standards User-Interface Design Posting and Review Process Management of User Accounts SYSTEM-TEST DESIGN Assessing the Job-Posting Process Assessing the Search Engine User-Evaluation Survey FUTURE CONSIDERATIONS...23 APPENDIX A: CREATESURVEY QUESTIONS AND FORMAT...24 APPENDIX B: CREATES URVEY.COM RESULTS (COMBINED ALUMNI AND STUDENT SURVEYS)...25 APPENDIX C: ALUMNI SURVEY RESULTS...33 APPENDIX D: STUDENT SURVEY RESULTS...39 APPENDIX E: STUDENT INTERVIEWS...43 APPENDIX F: FACULTY INTERVIEWS...47 APPENDIX G: CAREER SERVICES INTERVIEW...49 APPENDIX H: INFORMATION TECHNOLOGY (IT) STAFF INTERVIEWS...52 APPENDIX I: WEB SITE EVALUATIONS...53 iii

4 APPENDIX J: BOARD MENU HTML...55 APPENDIX K: GEOGRAPHIC REGIONS OF THE UNITED STATES...57 iv

5 1. EXECUTIVE SUMMARY This report presents the results of a systems -analysis study of the process of making professional job announcements available to students and alumni of the Graduate School of Library and Information Science (GSLIS), The University of Texas at Austin. This process is part of the activities of the GSLIS Career Services Office, whose director is Mr. Ronald D. Pollock. The study was undertaken during June and July 2000 by the students in course LIS 387.5, Systems Analysis and Evaluation, taught by Dr. Ronald E. Wyllys. Mr. Pollock presented Dr. Wyllys and his systems -course students with the challenge of improving the job-posting process and design. 1.1 SUMMARY OF IDENTIFIED REQUIREMENTS The need for improvement stems from several factors. Among these are issues that relate to both administrative staff concerns and end-user serviceability. The Systems Analysis Team (SAT) distributed among themselves the tasks necessary for analyzing and recommending a solution for improving the job-posting process and resulting information that is provided to students. The tasks performed included conducting surveys, both personal and via , of students, alumni, Career Services (CS) staff, and faculty. Through these surveys and through the information presented by Ron Pollock, the team was able to ascertain and define the requirements for the new, improved system. Requirements for the new system were found to include: Provide copy-and-paste functionality Permit full job description Efficiently avoid duplicate information Provide daily updates List job information by category Allow for keyword search Limit access to authorized users Mr. Pollock had developed a test Internet site using Discus, a bulletin-board software application available to the CS office. Since monetary constraints are a consideration, the Discus solution was chosen as the preferred alternative, and was consequently the focus of the team s evaluation and recommendations. 1.2 SUMMARY OF RECOMMENDATIONS On the basis of the identified requirements, the SAT developed a proposed solution. First, Discus was measured against the requirements to ensure that critical functionality would be met. Although Discus provides the necessary functionality required for the job-posting solutions, some limitations were identified. The two most apparent limitations were that Discus does not provide any validation rules or data-checking mechanisms. Second and more important, the search capabilities are minimal, since there are no fields of data by which to search (as there would be in, for example, a relational database). The design proposed within this document attempts to address all identified requirements, as well as the limitations of Discus. Discus, a software developed for message-board usage, is proposed to be used display job-postings in the place of messages. These job-postings should be categorized by job type, and then subcategorized by location. To further aid the user in easily browsing and searching for job postings, we recommend that each job include a red title that will enhance the visual organization of information. This report outlines the process for posting jobs via Discus and includes a copy-and-paste methodology that should greatly reduce the amount of time required for entering job postings. In additional, duplication of entries should be limited by utilizing the search or the browsing capabilities of Discus. Through standardization of job titles, any administrative user should be able to locate unique job postings. User-account maintenance and archiving features are also addressed as applied to the Discus environment. Once the Discus job-posting site has been designed and implemented, we recommend that evaluations be made regarding user satisfaction and process efficiency. The criteria and methodology for assessing the job-posting process, Discus search engine, and user evaluations are outlined in Section

6 2. INTRODUCTION 2.1 MISSION STATEMENT The project team will analyze and make recommendations to improve the system for posting job notices through the Career Services Office at the Graduate School of Library and Information Science, The University of Texas at Austin. The analysis will include reviewing historical and current procedures and the system used for posting job notices. The overarching goals of the reengineering will include reducing the cycle time for posting jobs, minimizing staff-training requirements, increasing system availability, and improving user satisfaction and usability. 2.2 OVERVIEW This document was created and generated by GSLIS graduate students. The team of students identified itself as the Systems Analysis Team (SAT) and consisted of: Ms. Jie Chen, Ms. Amy Gannon, Ms. Yu -Lu Hsiung, Ms. Jia Le, Mr. Steven Schang, Ms. Judi Tidwell, and Ms. Stacey Wuest. The SAT performed the analysis under the direction of Dr. Ronald E. Wyllys in order to fulfill the requirements of LIS 387.5, Systems Analysis and Evaluation, in the Summer 2000 semester. After defining the mission statement, the SAT performed its analysis of the Career Services Office (CSO) jobposting process by distributing and completing the tasks of conducting surveys of students, alumni, Career Services staff, and faculty. Through these surveys, additional research, and information presented by Mr. Pollock, the team was able to ascertain and define the requirements for the new, improved system as defined in Section 3.4. The SAT has made recommendations based on findings. These recommendations define the configuration of Discus as a job-postings solution, the elements of the Career Services Office job-postings Website, and the process for posting and administrating the job announcements. These recommendations are found in Section 4.2. The team has also addressed system-test-design procedures and future considerations. 3. ANALYSIS 3.1 SYSTEM HISTORICAL DATA The Career Services Office maintains and publishes a weekly job-placement bulletin. This bulletin is a comprehensive list of available jobs in the field of Library and Information Science. In 1994 the current system to publish these job notices was started. Prior to this job notices were posted by hand on a bulletin board at the GSLIS. The new system in 1994 strove to automate the process and improve the service to students and alumni. The new system used Microsoft Access and Microsoft Word to manage job notices. The job notices were published weekly. The majority of subscribers received the job bulletin electronically via although several people received a printed version through U.S. mail. The initial focus of the Job Bulletin was on listing jobs available in Austin and Texas. As interest grew in the publication, there was a need to post national and international jobs. Furthermore, the increased use of the Internet made it possible for the Career Services office to find jobs outside of Texas. These two factors led to the creation of Weekly Placement Bulletin. This bulletin lists national and international jobs in Library and Information Science. Texas jobs were then distributed through the Texas Weekly Placement Bulletin. Since their inception the job bulletins have been a success. In 1999 the Career Services Office published over 9000 jobs and had a subscription list of over 280 people. The original listserv effectively disseminated new job postings to students and alumni. However, today the current job-posting system does not provide information in some of the way in which the users would prefer it. The system is also in need of administration improvements to facilitate Career Services staff requirements. 2

7 3.2 CURRENT CSO OPERATIONS AND ORGANIZATION The Career Services Office (CSO) currently has only two staff members who process all job-posting entries. They enter approximately job postings to the Access-driven listserv each week. The staff maintains the employment boards for the professional and student jobs as well as posting volunteer and internship opportunities. Along with entering the job opportunities, they have the added duties of searching out new positions and companies within the World -Wide Web and in person in our community. They are also in charge of organizing tutorials for interview techniques, résumé tips, and general job-hunting practices. Despite the list of duties expected from the Career Service Office staff, the priority goes to maintaining the job boards. Since the current system has such limitations for data-entry options, the staff spends the majority of their time just entering the supplied announcements. Currently, the first step in the posting process is in gathering the information. Once they have all of the job postings for that week, they have to manually sort them to eliminate the duplicates. Since the system is -based, each job-listing has to be entered each week. A disproportionate amount of the staff time is used in entering the data from the hard copy and submissions (up to twelve minutes per item). The system also limits the amount of information that can accompany any one posting, so the staff has to re-word many of the notices to supply as much of the contact and descriptive information as possible for the users. The listserv has nearly 280 subscribers, including both alumni and current students in the GSLIS program. The fees collected by the alumni who have been out of the program for over a year are used to supplement the basic functioning costs of the center. The budget for the CSO is taken from the student activity fund and, with such a small group of students contributing, the budget is highly limited. Consequently, the staff must actively maintain the status of alumni who are on the listserv and notify them when their dues are required. Currently the system is unable to sort the users, so the staff has to keep hard copies of the list and the graduates. The reliance on hard copies of old announcements, student lists, and current employer sources requires shuffle time for the staff that could be used in a more productive area. Figure Current Service Flow Chart Gather Job Announcements Print Hard-copy Of Job-listings Weed: Check for Duplicates Enter Job Data Post: Texas Jobs Weekly and Weekly Placement Bulletin 3

8 3.3 SYSTEM ENVIRONMENT FOR JOB POSTINGS Currently job announcements used for the Career Center employment board come to the office either through , postal mail, fax, or phone calls. When employers calls the office, they receive priority because there is usually a stronger time constraint on the position. The staff attempts to gather more employment opportunities from Websites they discover which may be looking for employees with the skills taught in the program. However, time restrictions limit this source of information. Since the data-entry portion of the process is so time-consuming, the staff has to be very careful about identifying duplicates and in sorting the job positions. The postings with the closest closing date are given next priority after the phone notification. However, if the position is likely to fill within two weeks, the staff will most likely reject that listing. The review process encompasses checking for accurate contact information, making certain that they are not duplicate notices, re-writing the announcement to match the parameters of the current system page format, sorting for job type and location and the time limitations for the applications. None of these processes can be handled electronically with the current system. 3.4 SURVEYS Survey Results The survey was chosen as a way to measure the preferences of both the current students and the alumni who are active users of the GSLIS Professional Jobs Listserv. Since the current system of the Professional Jobs Bulletin Listserv is -based, it easily targeted all of the current users. Out of the approximately 280 subscribers, 82 subscribers participated in this survey. The survey utilized both open-ended and closed-ended question formats. The closed questions used a combo-box with 2-3 choices per question. The closed questions allowed the users to give us their unique feedback. Both types of questions will help to create a system that meets the current needs and desires of users of the Professional Job Bulletin Listserv. A copy of the survey can be found in Appendix A. The survey was produced with CreateSurvey.com creates and analyzes user-defined survey data. The designed survey was created, submitted to Mr. Pollock, and ed by him to all of the current subscribers. Complete results (a compilation of alumni and current students) from CreateSurvey can be found in Appendix B Alumni Survey Results Complete alumni results can be found in Appendix C. Closed-Ended Question Results Do or did you utilize the GSLSI professional jobs bulletin? 94% utilize the Professional Job Bulletin Listserv Did you accept a job listed by the GSLIS professional job bulletin? 54% have accepted jobs listed by the listserv During our job search, did you search primarily by location or job description? 33% search primarily by job description 57% search primarily by job location 10% search by both job description and location 4

9 Open-Ended Question Results By what other categories do or would you like to search in job-postings? Some of the most popular results are: 32% Salary 16% Type of Job (Librarian, Information Technology, Archivist, etc.) What aspect of the GSLIS Career Services did you find MOST valuable? Results centered on the following: 20% Regularity of postings 20% Timeliness of postings 38% Comprehensiveness of listings (esp. in Texas) What aspect of the GSLIS Career Services did you find LEAST valuable? Results centered on the following: 8% Not comprehensive for locations other than Texas. 14% Many jobs not related to the library field Current-Student Survey Results Complete student results can be found in Appendix D. Closed-Ended Question Results Did you accept a job listed by the GSLIS professional job bulletin? 29% Have accepted jobs listed by the listserv Do or did you utilize the GSLSI professional jobs bulletin? 65% Utilize the Professional Job Bulletin Listserv During your job search, did you search primarily by location or job description? 21% Search primarily by job description 44% Search primarily by job location 35% Search by both job description and location Open-Ended Question Results By what other categories do or would you like to search in job-postings? Some of the most popular results are: 37% Salary 11% Type of Library What aspect of the GSLIS Career Services did you find MOST valuable? 11% ed jobs What aspect of the GSLIS Career Services did you find LEAST valuable? 16% Lack of salary listings Student Interviews The overall purpose of the new professional job-list service is to satisfy the users. To find out the students (users) expectation of the service, we interviewed some current GSLIS students who actively use the current system. These interviews determine preferences for the professional job-list service, and suggestions for improving the current system. First, the student-name list was obtained from Mr. Pollock and from Ms. Carol Carreón, the GSLIS Graduate Coordinator. The names, phone numbers, and addresses of the students who are quite active in job searching were provided. Many of these students have been to the CSO previously for information regarding job searching. 5

10 According to this selection criterion, selected interviewees were fairly representative of all the GSLIS students who could use the service. After making the selections, we determined questions and methods for conducting the interviews by consolidating separately generated lists of possible questions. We interviewed the students either by phone or in person. A total of 15 interviews were conducted: 8 via phone and 7 in person. Finally, we analyzed the interview results and summarized them into a table. See Appendix B for student-interview questions and the table of interview results. Do you know about the listserv? How did you hear about it? (If not, skip to #7.) 10 out of 14 students we interviewed had heard of the list service. They got to know this service either from Ron Pollock s presentation, or from the GSLIS receptionist, or from their friends. Have you used the listserv? How often? 7 out of 14 have used it; 5 of the 7 users use it very often. How would you rate the listserv according to your overall satisfaction with the service? Excellent 4 Very good 1 Good 1 Fair 1 Poor 0 What is one thing that the listserv is doing great and one thing that needs to be changed? Good: (1) Frequency of updating, (2) enormous number of job-posting in the list, and (1) publicity. Change: (7) Need more descriptions, (3) easier to read, and (3) only see positions of interest. Did you actually get an interview as a result of using the listserv? 3 out of 7 students did get interviews; one got two interviews. Is it the main channel you use for job searching? 5 out of 7 use the list service as the main channel. What (other) services or channels do you use for job hunting? 3 out of 13 Newspaper or chronicles 9 out of 13 Other job-search Web sites 3 out of 13 Librarian associations What categories do you prefer to use for job search? (Keyword, description, location, salary, etc.) Note: Respondents may have more than one preference. Location 9 students prefer to use location as their criterion. Description 7 students prefer to use description as their criterion. Salary 3 students prefer to use salary as their criterion. How much are you willing to pay for a service like this? 5 out of 14 students not willing to pay for anything. 4 out of 14 students' willingness depends on whether the interview is successful. Others willing to pay $5 - $10 per month or $15, $25, or $100 per year. Do you have access to the Web to use the new service? 14 out of 14 students have access. 6

11 Following is a synopsis of what students expect the new job-posting service to provide them: The new listserv should be able to help students to use other job-searching channels, and to provide instant links to other Web sites. Students need other channels (like other Web sites, newspaper, announcements of some company etc.) to conduct job searches. When students search for jobs, they usually start from one Web page and go to other related pages through links. Adding links to the CSO job board would add convenience and save time. For example, one of the links a student suggested is University Human Resources. So that we may know what s going on there one of our interviewed students remarked. Students also want links to other job-searching engines and other resources links for job seekers. The new system should be able to provide a searching engine. Students searching for jobs are concerned about the salary and location of the job. For instance, 11 out of 15 students interviewed want the new professional job-listing service to provide information about salary and/or location of the job. As there should be a wealth of information provided for the students, it is impossible to scan all the listed jobs just to get one or two. One of the best solutions to this according to our interview is to set up a keyword search. Some 11 out of 15 students preferred to use location and salary as their main category, and 8 out of 15 students preferred to use job description (or job title) as the main category. They also want the information to be categorized in an orderly fashion, and they want the new search engine to be able to provide them cross-category searches. For example, when they type Austin, library, $50,000 the new listserv can give them links that contain all these words. The new system should be able to provide a full description of the job. Once students find a potential job, they want to know more information: for example, detailed descriptions (including salary) and requirements for the job. Also of interest to the students is the company environment, culture, and philosophy. A link to the company Website would be helpful, if available, and any contact information. In addition, they are very interested in the salary and all the other benefits that the company offers. Tips for interviews and job searching. Among the 15 students interviewed, 3 students wanted the new listserv to provide them ideas about the interview and job-searching processes. For instance, what should they pay attention to when having an interview? What should they know before the interview? Besides the above, students also want suggested resources for learning more about listed jobs through the new job-posting board. Easier interface to submit and modify résumés. When students go to the professional-job listserv, they usually want the list to be easier to read. They also prefer to see only the jobs relevant to them. More non-library jobs. Since not all of the students want to be a librarian after graduation, interest in non-library jobs exists. Among the 15 students we interviewed, two of them remarked that they want some more non-library jobs to be posted on the new job-posting board Faculty Interviews In order to collect from faculty opinions of the current Career Services Office (CSO) job-posting system, four faculty members in GSLIS were randomly chosen and interviewed. Interview questions were provided in advance to all the chosen candidates through . One response was provided directly through because of schedule conflicts, while the rest were conducted face-to-face. Each interview took approximately 10 minutes for the following four questions: (See Appendix F for detailed information.) What do you like and/or dislike about the current Career Services Office job-posting listserv? What features would be useful to have on a job-listing system? What recommendations do you give to students who are job searching? Do you currently subscribe to the Career Services Office listserv? All of them think the CSO job-posting service system is useful and provides a valuable job-searching tool to students and alumni. Although only one of the four subscribes to the Weekly Job-posting listserv, all of them are familiar with it. They all forward individually-sourced job information from academics or companies to CSO immediately after receipt. One of them also consults the site occasionally. 7

12 When asked, What features would be useful to have on the job-listing system?, respondents emphasized listing job information by categories, which they believe will help our students and alumni to seek positions by type, rather than having to search through the whole database to find the one that fits their needs. They felt that substantial amount of time and effort will be saved via the categorizing of information. They also mentioned that there is a great need to translate the words used to describe the jobs and careers into conventional terminology, so that students and employers have a reasonably standard lexicon for dealing with the variety of descriptions. In the current system, although the functions are consistent, the descriptors used to describe functional job requirements change very rapidly. Two of them also mentioned the importance of giving full job descriptions, as this will help our students and alumni more efficiently prepare their résumés and communicate with the employers. The recommendations to students who are job hunting given by four of our professors are: Prepare an outstanding résumé. Relevant work experience is extremely important, yet good academic standing is also an important factor Try to gather as much job information as possible. Use the CSO job-posting as a major source, but also make use of other available job resources such as many on-line job databases provided by a variety of companies, newspapers, bulletin boards, etc. Speak with alumni or people recently employed in field. Contacting people who are working or just went through the job-hunting experience in relevant positions is a shortcut to gather first-hand information. Be persistent and don t lose hope at any time. Job-hunting can be a long and difficult process. Be mentally prepared and don t get frustrated by any unsuccessful experience. To maximize the chance, don t stick all your hopes on one or two positions Career Services Interviews The purpose of interviewing with the CSO was to identify the deficiencies and problems of the current job-posting system, and also to address the concerns of the CSO staff. Valuable first-hand information was obtained through personal interviews about the most needed improvements and possible ways to achieve these improvements. The director of the CSO, Mr. Pollock, and the two staff members, Ms. Grace Elinsway and Mr. Ismael R. Archbold were interviewed. Each interview lasted one and half hours. Interview questions were divided into three categories: Current System (14 questions) Desired System (7 questions) Other Information Affecting Analysis (5 questions) Refer to Appendix G: Career Services Office Interview for complete set of interview questions Current Status Overview From the interviews it was determined that the current job-posting system (listserv) is a very useful job-searching tool for GSLIS students and alumni, and is being utilized readily. For example, in 1999, a total of 9113 jobs were posted to the system, which averages 759 monthly, 178 weekly, and 35 daily (these figures are based on the 51- week publishing schedule). Currently over 280 people are on the subscription list, and the CSO staff is working as efficiently as possible at searching and posting jobs using limited resources. They often find job information through the Internet or other means prior to submission from the contributing organizations. The job information is submitted via a variety of formats including , printed mail, fax, classified advertisements, and bulletins from Internet job sites. The job-postings originate from a variety of sources, which include professors, alumni, business people, trade magazines, academic institutions, companies, and the search efforts of the CSO staff Deficiencies and Addressed Concerns During the course of the interviewing process, inquiries were made as to the shortcomings of the listserv process and possible improvements to eliminate these shortcomings. 8

13 The first problem identified is that the listserv derives its information from a Microsoft Access database. The way the database is designed, all of the job information has to be manually typed into the database, regardless of the format. It takes an average of 12 ½ minutes for each position to be entered. This process is so time-consuming that it greatly limits the job-collection volume and reduces the information-update rate (the job bulletin is updated once each week at this time). In addition, there is not enough space to store detailed full job descriptions that may include application processes and position responsibilities. Currently, CSO can only provide an abbreviated format for the above job information. This makes it difficult for job seekers to prepare their résumé and cover letters efficiently. CSO wants to find a way to solve these problems. They want to install a new software system (Discus), which both supports a cut-and-paste function to eliminate the labor of manual operations, and also allocates enough space to permit storing full position descriptions. The desired input time with the new software is approximately two minutes to input each job announcement. If this was achieved, the saved time could be used to search for more job-listings in a greater scope to provide better service to GSLIS students and alumni. Also, with the reduced time per job entry, information could be updated on a daily basis, and give customers around-the-clock accessibility. Another major factor identified is the time-consuming process spent on preventing duplicate entries. Duplicates are a result of companies or libraries sending job information to CSO more than once or in more than one format. Students, alumni, and faculty also send job positions to CSO through . When a new job is received, a query for all jobs from that organization is performed. The result of that query is compared to the new job to ensure that it is unique. Currently in order to do so, multiple time -consuming steps are required (listed below). The CSO staff is eager to find an alternate way that requires less time and work to solve this problem. - Place cursor in field you want to search. - Go to toolbar and click on Search. - Click on Find - Enter words you want to search for. - Check boxes for search Fields as Formatted and search only Current Field. - Select search up. - When next record pops up (if there is another entry for that job) select search dropdown menu, select Filter by Selection (can further refine the filter action). - When finished, select remove filter/sort from search dropdown menu. According to CSO staff, students are increasingly focused on job type rather than job location. Therefore, the CSO wants to improve the way through which job information is posted. They plan to post jobs by type (including academic library, public library, school library, archives management, special library, and Web and/or Internet) and location (state listings). This way, students and alumni can quickly locate the job that fits their interest rather than browsing through the whole database, thus saving a substantial amount of time in the job searching process. The CSO also wants to provide users with keyword-searching capability, such as salary, job type, location, closing date, etc. Eventually, the CSO staff would like to provide the combined-keyword search to enable highly efficient information categorizing. Additionally, Web-based bulletins are planned which would have links to source institutions and/or companies for making navigation easier among different scopes. Another problem with the listserv addressed by CSO is the difficulty in tracking subscriptions. Current students and alumni who left school less than one year ago can use listserv for free, while those who left more a year ago need to pay a usage fee for the service. As a result, CSO staff must continuously trace the free and non-free subscriber information. This process is entirely manual and error-prone. They hope the future Web-based posting system can ease this problem by tracking the status of students and alumni, and by automatically performing the switching between free and non-free subscriber lists. Money is always a factor. The CSO annual budget is limited, with its main source from student-activity fees. Since GSLIS has only a small group of students, the amount is restricted (there is only $700 available for system improvements each year). Therefore, any new software has to be at a minimum cost. In addition, it should be easy to use in order to limit training and technical skill requirements for staff. 9

14 Finally, the CSO wants to eliminate the response comment feature on the new system to avoid users trying to post jobs or comments without CSO approval Summary In conclusion, the Career Service Office is eager to help create a better job-posting environment for the students and alumni as well as for themselves. To satisfy this purpose, they would like the new system to possess the following features: Support copy-and-paste function Permit full job description Efficiently avoid duplicate information Permit daily updates List job information by categories Allow for keyword searching Limit access to only authorized users Be cost-effective Be able to get feedback from users Desired Capabilities The data collected from the previous sections were studied and analyzed by the systems -analysis team. All desired requirements were condensed into the following list: 1. The system will provide for each job posting a full, detailed job-position description and context, applicant qualification, and full information on résumé requirements. 2. The system will require of the CSO staff less than four minutes to process and post each job position. This short processing time will make daily update of the postings feasible. 3. The system will eliminate the need for entering job information by hand. All job positions will be posted by copy-and-paste functions. 4. The system will be Web-based. It will permit user access at all times, twenty-four hours a day and seven days a week. It will allow employer to submit job-opening information on the CSO Web page. 5. The system will have limit access to only authorized users. 6. The system will allow the users to perform keyword searching by location, job title, salary, etc., by a full-text search, and by a title and/or heading search. 7. The system will allow the users to browse the Web site by job title, location, etc. 8. The system will provide a channel for users feedback on the job-posting service quality during the system evaluation phase. 9. The system will provide hyperlinks within the Web site and to the employer Web site and other job-search Web sites. 10. The system will be able to store job positions after one month in an organized archive. 11. The system will assure all job postings have standardized headlines. 12. The system will be easy to use and will be implemented with minimal training for system administrators. 10

15 4. RECOMMENDATIONS 4.1 DISCUS AS A JOB-POSTINGS SOLUTION Discus and Identified System Requirements The following is a list of the Desired Capabilities as defined in the Analysis Phase compared to the Identified System Requirements for Discus. 1. The system will provide for each job posting a full detailed job position description and context, applicant qualification and provide full information on résumé requirements. Discus DOES fulfill this requirement by providing ample space to include complete descriptions. 2. The system will require of the CSO staff less than four minutes to process and post each job position. This short time for processing will make daily update of the postings feasible. Discus DOES fulfill this requirement by enabling an average of approximately two and one half minutes per posting. 3. The system will eliminate the need for entering job information by hand. All job positions will be posted by copy and paste functions. Discus DOES fulfill this requirement by allowing cut and paste entries. 4. The system will be Web-based. It will permit user access at all time, twenty-four hours a day and seven days a week. It will allow employer to submit job opening information on the CSO Web page. Discus DOES fulfill this requirement by enabling the CSO to place job-postings on the Internet, thus providing users to access information at all times. 5. The system will have controlled access to authorized users. Discus DOES fulfill this requirement by supporting private reading topics (password required to read joblistings). 6. The system will allow the users to perform keywords searching by location, job title, salary, etc., by a full text search, or by a title and/or heading search. Discus DOES fulfill this requirement by featuring keyword searches. 7. The system will allow the users to browse the Web site by job title, location, etc. Discus DOES fulfill this requirement through configuration in a variety of ways, including how the jobs are posted. (For example jobs can be posted by job title, location, etc.) Also a keyword search allows the user to search by job title and location. 8. The system will provide a channel for users feedback on the job-posting service quality during the system evaluation phase. Discus DOES fulfill this requirement through the availability of a user feedback function. 9. The system will provide hyperlinks within the Web site and to the employer Web site and other job-search Web sites. Discus DOES fulfill this requirement because hyperlinks are supported. 10. The system will be able to store job positions older than one month in an organized archive. Discus DOES fulfill this requirement by offering two ways to archive: automatic archiving and pruning by conversation length and manual archiving and pruning by date. 11. The system will assure that all job postings have standardized headlines. Discus DOES fulfill this requirement if configured to create standardized headlines. 11

16 12. The system will be easy to use and will be implemented with minimal training for system administrators. Discus DOES fulfill this requirement, being a very user-friendly program, which enables it to be learned quickly and easily by system administrators Limitations of Discus Discus is full-featured and powerful software. Although it can be adapted for use as a job-posting system, its original design was as a bulletin-board system. These two functions, bulletin board and job postings, have many similarities, but there are some notable differences. These differences will equate into limitations of using Discus for posting job notices. This section will highlight those limitations. Also, it will point out areas where Discus will be effective as long as the processes providing input to Discus remain sound Information Validation The method for entering a job notice onto Discus is to paste free form-text into a text area on an HTML form. This form is submitted, and the text is appended to the job-posting board. The only information that Discus is forced to post is the name of the poster and date the message was added. There is no function to guarantee that the jobposting includes essential data (i.e., detailed position description, résumé requirements, contact information, etc.). For example, blank messages could be posted. Discus would not provide any kind of field checking or validation to prevent this sort of occurrence. Checking the validity and completeness of the content will be left entirely up to the person entering the information Searching Keyword searching poses a problem in Discus. A situation could arise where returned information will meet the search-engine criteria but might not necessarily be the information the user wants. An example would be a jobposting that includes a company s home-office address in Texas, while the position offered is in New Jersey. The user searching for jobs in Texas does a keyword search. This job posting will be returned because it contains the word Texas. The end result is that this would be an invalid search result for the user because the job is in New Jersey, not in Texas. The problem is caused by a lack of discrete fields in any given job posting. A job posting is essentially one long string of text, which Discus searches for keywords. This is contrary to how a database could be designed, where a job posting would be a record divided into several fields. These fields could contain specialized information like job location, employer, or salary. Each field could be independently searched, effectively narrowing search results to match user requests. See Section for recommendations regarding the configuration of Discus to alleviate the above-mentioned shortcomings Alternative Solutions Using Discus as a system for posting job announcements meets the identified system requirements in several ways, but also presents administrators and end users with limitations. In order to address these limitations, a database solution might be considered as an alternative solution. The CSO currently uses a Microsoft Access database to store job postings. However, the database is programmed in a way that limits the amount of data that can be stored, and forces users to type data manually as opposed to a cut-and-paste method. By redesigning or creating a new database, administrators could cut-and-paste job announcement information into a database form, which would store the data in the tables. Utilizing MySQL and PHP, (or an equivalent technology compatible with the current server configuration), linking a job-postings database to the Career Services Web page is possible. With a properly designed user interface, end users could query the database for job-postings based on virtually limitless criteria. In addition, administrators could easily ascertain what jobs should be archived or re moved from the database through the use of queries. Duplicate entries could be identified at the time of posting by creating a duplicate-field primary key in the database, which could use the job title, company name, and location combined as a unique identifier. This unique identifier would prevent duplicate entries. 12

17 The most apparent drawback to this solution is the requirement for database programming and Web-interface design. Maintenance of the database would also be a factor to consider. However, the functionality of using a database solution would provide end users with job-postings that meet all identified requirements. See Appendix I for job-board evaluation listings. The majority, and the most functional, of examples of other job-posting sites evaluated appear to be driven by a database solution. End users can perform complex queries by multiple criteria, and are presented with summary or full description options. Given the limited search capabilities available with Discus, an active database with a Web interface may be entertained as a possible consideration, dependent upon available programming and maintenance resources. Alternatively, if the CSO wishes to incorporate a more sophisticated version of the bulletin-board solution, Discus Pro provides the following added features that would be applicable to the present solution: User self-registration Manual archiving and pruning by date Attachment uploading Cookie-based login notification enhancements Preferences of users stored to control board options Although these features are desirable, they are not necessary to meet the identified system requirements. 4.2 PROPOSED DISCUS DESIGN Job Categories and Title Standards In order to reduce some ambiguity in search results as identified as a limitation of Discus in Section , subcategorizing the job boards by job type and then geographical region would help users sort search results. Discus outputs the name of the board and a link to the posting for a search. Hence, the boards should follow the convention of Job Type > Geographical Location > Job postings. The user could then scan the search results for job type and location and select the appropriate postings. Although the basic divisions used in the current system could be maintained, the proposed Discus configuration supports additional categories that would improve on user searches. The categories recommended are as follows: (in alphabetical order) Academic Administration Academic Archives Academic Reference Academic General Technical Services Faculty Positions Learning Resource Information Science Miscellaneous Public Library General Public Library Reference Public Library Youth and/or Children The job listings would be organized into these categories, and then subdivided into regions. The regions will just be classed as: Texas US Regional (Subdivided by NorthEast, SouthEast, MidWest, SouthWest and West. See Appendix K for state regions.) 13

18 International The second feature to reduce search-result ambiguity would be to make the first line of the job posting a standardized title. The standardized title would follow the format: position title, company, and location. For example, Systems Analyst III, Motorola, Austin TX. When Discus returns the results of a search, it automatically displays the author, date posted, and the first line of text in the posting. If the first line of text in the posting is the standardized title, then users could scan the results for relevant information. To further ease navigation through search results, we recommend that the title be posted in red so that the user can readily identify the beginnings of postings. See Section for instructions User-Interface Design The user interface (Figure ) should have an intuitive design that will support primary user requirements. There should be hierarchical and lateral navigation structures. Navigation structures will promote both logical searching and browsing for job postings. A method to indicate user position in the Website should be implemented. The primary users are defined as students and alumni job seekers. Figure Discus Homepage 14

19 Navigation The left frame (Figure ) on the page would be for top-level navigation, searching, documentation, and feedback (if desired). The left frame would link to the top-down hierarchical navigation through the tree view link. Also it would provide top-down browsing from anywhere in the site by allowing the user to select categories. This navigation bar would be the primary point to start navigating the job postings. Links from this navigation bar would load pages in the right frame. This navigation bar would remain in view during all navigation in the site. Links for feedback, search, and documentation would be conveniently displayed in the left frame. Figure Left Frame The right frame (Figure ) would provide site locations and the job postings. The site location would be a list, separated by colons, :, of page titles at the top of the frame. The list is read from right to left. The current page 15

20 would be the right-most link; going left would be the next higher page in the site hierarchy. This arrangement would allow the user to maintain situational awareness while navigating and would provide a way to back out of the current page. Job postings would be in a list, with the most recent at the top of the page and the oldest posting on the bottom of the page. Figure Right Frame Job postings would be lis ted with a red title (Figure ). The red title should summarize the job title, company, and the location of the job. The color would allow for quick scrolling between jobs when there is a long list of jobs in the right frame. The format of the red title would allow the user quickly to identify key aspects of the job. This title bar also creates a summary line when a search is performed and a list of job-postings is returned. 16

21 Figure Sample Posting Posting and Review Process Information Verification When a piece of job information is received, regardless of format ( , printed mail, faxes, classified advertisements, or bulletins from Internet sites), it should first be checked for completeness. Postings should contain as much information as possible, to include: job type, job title, position responsibilities, minimum requirements, desirable qualifications, application process, required materials, salary, closing date, and contact information. Occasionally, some information may not be available, such as salary or desired qualifications. However, no critical information such as the job provider s name and contact information should be missing. If such a situation were to occur, the job provider should be re-contacted for the missing items before posting it in the system Avoiding Duplicates After information verification, the second step should be to prevent duplicate information. How to avoid this redundancy efficiently is one of the biggest concerns of the CSO, as checking duplicates consumes a lot of human 17