Systems Analysis and Design

Book: Systems Analysis and Design 6e Kendell & Kendall

Chapter 4 Information Gathering: Interactive Methods

1 Interviewing
The first step in interviewing is to discover where your own biases are. Interviews are in the question and answer format. But you do need to get the feelings of the interviewee. Seek their opinion. They know the organization better than you do. They have the information needed to make the project success. The interviewee is probably a stranger to you, so you will need to build up trust with them. This is done by preparation.
A. Five Steps in Interview Preparation
I. Read background material.
Check web sites, annual reports, etc. to get information on the organization.
II. Establishing Interviewing Objectives
Use the information to establish objectives.
III. Decide Whom to Interview
Strive for balance in whom you interview. Do it at many levels in the organizations.
IV. Prepare the Interviewee
Call or e-mail the person to prepare them. If detailed interview, send them questions to prepare.
V. Decide on Question Types and Structures
Decide on what type of questions and type of interview.
B. Question Types
I. Open-Ended Questions
Responses can be two words or two hundred. The interviewee responds how they feel best. Benefits include; interviewee is put at ease, interviewer can pick up jargon of interviewee, more interesting for the interviewee, more spontaneity, if caught unprepared it is a good technique. Drawbacks include; answers may contain too much and irrelevant information, chance of losing control of interview, looks like interviewer is unprepared, looks like interviewer is on 'fishing expedition'.
II. Closed Questions
The questions are answered with detailed facts. It limits the ability of the interviewee to respond. Special type is bipolar question. This is a yes or no type of question. Benefits include; saving time, ease in comparing interviews, control of interview, gets to relevant data. Drawbacks include; boring for the interviewee, missing the rich detail, no rapport built.
III. Third type is a probe question. You get the interviewee to clarify earlier question. Do not be afraid to use this question.
C. Arranging Questions in a Logical Sequence
I. Using a Pyramid Structure
Start with specific questions and go to the more general questions. Use it if interviewee needs to warm up to you
II. Using a Funnel Structure
Starts with the more open ended and ends with the specific ones. Good for the interviewee needs to have the freedom to express things because of emotional nature of interview.
III. Using a Diamond-Shaped Structure
Combines the two, starts with pyramid to get things go with closed questions. Works its way to pen questions and then starts the funnel and narrows them down to closed questions.
D. Writing the Interview Report
After you are done write up your notes on the interview. Type to review the report with the interviewee to make sure you got things correct and to show you are interested in their opinion.
2. Joint Application Design
Developed by IBM JAD is used when time (and money) is of the essence. It requires proper training to be able to run a session.
A. Conditions that support the use of JAD
I. Users want something new not standard answers
II. Organizational culture supports joint-problem solving behaviors across the organization
III. One on one interviews will not generate enough ideas
IV. Workflow allows for absence of key employees
B. Who is Involved
Several people will be involved. One IS analyst should be in the group. You as the project analyst will be there. Eight to a dozen people should be involved, including analysts, users and executives. They should be above clerical level. One scribe should be there from IS department.
C. Where to Hold JAD Meetings
They should not be held in the office but someplace where there will be few interruptions. It should only be held when all participants are available. If possible do a pre meeting about a week ahead so that everyone will know what is expected of them.
D. Accomplishing a Structured Analysis of Project Activities
All the questions brought forth should be asked the five Ws: Who, what, where, why and how.
E. Potential Benefits of Using JAD in Place of Traditional Interviewing
I. Time savings
II. Rapid development
III. Improves ownership of information system
IV. Creative development.
F. Potential Drawbacks of Using JAD
I. Large block of time commitment
II. If preparation is inadequate, or if follow up not done properly, results will be unsatisfactory
III. Organization may not be mature enough to do a JAD
3. Using Questionnaires
When people are wide spread or need to do things in their own time, a survey or questionnaire might be a good tool. In using those questions must generally be closed in nature if quick turn around time is desired.
1. Planning for the Use of Questionnaires
A proper questionnaire can take some time to develop. You must decide on what you are trying to find out. Once that is done decide on the type of questions
2. Writing Questions
Questions need to be really clear since interaction is generally impossible. Make sure that your questions if they are open ended are not too broad. Anticipate what they will answer. Use closed question whey you are able to list the items that they could respond with. One also has to be careful with the choice of words on a questionnaire. Make sure to use the language of the company (do they call them managers or supervisors). If necessary ask a test group to help you out with the correct wording.
3. Using Scales in Questionnaires
I Measurement - there are two different ways to scale your answers, nominal scales and interval scales. Nominal gives you a list and you pick from it. You then tally up the numbers for each item. Interval scales let the respondent chose from one extreme to the other, usually with a number scale.
II. Validity and Reliability - Make sure that the questions measure what they are intended to measure. If they do they are valid surveys. Reliability deals with consistency. If you took the survey again you should get the same results.
III. Constructing Scales - Problems are of three types
a) Leniency - easy raters which will move the average off of where it should be. This can be counteracted by moving the mid point to left or right of center.
b) Central Tendency - Everyone rates things as average.
c) Halo Effect - people take impressions formed and carry them on to the next question.
4. Designing the Questionnaires
Make sure there is enough white space, and allow ample room to write responses. Make it easy to mark their answers and be consistent in style. Cluster items of similar content together as well as put stuff important to respondent at the beginning. Controversial stuff should be at the end.
5. Administering Questionnaires
I. Respondents - Decide who will get the questionnaire. Make sure there is enough of a sample to be a fair representation. Remember many will not return the forms.
II. Methods of Administering Questionnaire - One should pick best way to distribute questionnaires. Most common is to let them self administer it, though this results in lower response rates. It is used the most because it allows the anonymity that many people want. Another method is to do it by email or web site.