Information Gathering Techniques

1
Information Gathering Techniques 1. Interviews 2.
Questionnaires 3. Observation 4. Reading
2
1. Interviewing Kinds of information sought
in an interview include - opinions, goals,
feelings and informal procedures By seeking
opinions rather than facts, SA can often discover
a key problem the interviewee wishes to
address SA can learn a lot from feelings
expressed by interviewees (who know the
organisation better) Interviews can tell the
SA more about personal and organisational goals
than other data-gathering methods
3
Planning the interview The five steps in
preparing for an interview are 1 Read
background information 2 Establish interview
objectives 3 Decide who to interview 4 Prepare
the interviewee 5 Decide what questions to ask
and how to structure the interview Possible
sources of background information include - a
current annual report, a company newsletter,
company information held in a public library
. The interviewee should be given time to think
about the interview and told how long it will
take, which should not be more than an hour
4
Establishing interview objectives Objectives
will depend on background information
gathered Aspects of information processing and
decision-making behaviour about which the SA may
want to ask questions include information
sources information formats quality of
information decision-making frequency decision-mak
ing style Types of interview question . The
three main types are open-ended questions, closed
questions and probes An open-ended question is
one that leaves open the interviewee's options
for responding e.g.
5
Will you please explain how you make a scheduling
decision? What is your opinion of the present
computer system? Open-ended questions have
several advantages 1. putting the interviewee
at ease 2. providing richness of detail 3. making
the interview more interesting for the
interviewee 4. revealing possible avenues for
further questioning 5. allowing more
spontaneity 6. filling a gap when the interviewer
is caught unprepared They also have several
drawbacks 1. getting too much irrelevant
detail 2. possibly losing control of the
interview 3. wasting time 4. giving the
impression of being unprepared or 'fishing'
6
A closed question is one that limits the
responses the interviewee can give e.g. How many
reports do you generate in a month.? Which of
these sources of information is most valuable to
you ? Do you use a microcomputer? Advantages of
closed questions include 1. saving
time 2. simplifying comparison of
interviews 3. getting to the point 4. keeping
control Their disadvantages include 1. boring
the interviewee 2. failing to build
rapport 3. missing important points
7
A probe is a question which is intended to
clarify or expand a previous answer
e.g. Why? Will you elaborate that point for
me? What makes you feel that way? Probing is
often essential to obtain full responses rather
than accepting superficial answers Rather than
seeming impolite, it can show that the SA is
listening and trying to understand the
interviewee's responses Question pitfalls Two
types of question which should be avoided are
leading questions and double-barrelled
questions A leading question is one that
suggests the response the interviewer would like
(or expects) to hear e.g.
8
You agree with other managers that stock control
should be computerised, don't you? The problem
with leading questions is that they are likely to
produce biased responses e.g. the above question
makes it uncomfortable for the interviewee to
disagree A much better question would
be What do you think about computerising stock
control? A double-barrelled question is really
two questions combined in one e.g. What
decisions do you make in a typical day and how do
you make them? Instead, the two questions should
be separated e.g. What decisions do you make in
a typical day? How do you make your decisions?
9
Double-barrelled questions should be avoided
because - the interviewee may decide to answer
only one question - the SA may misunderstand
which question is being answered and draw the
wrong conclusion - time may be wasted clearing
up any misunderstanding caused Advantages and
drawbacks of Interviewing in comparison with
other fact-finding techniques Four of the most
common methods of fact-finding used by Systems
Analysts are interviewing, questionnaires,
observation and reading/research. Interviewing
is commonly held to be the most effective form of
fact-finding, providing in-depth, valid
information.
10
Flexibility Its major strength is that being
conducted on a one-to-one basis, interviews
provide a situation where the researcher can
conduct the interview in a flexible manner in
response to the answers he receives. It is
possible to digress from the main area of
investigation if the respondent mentions
something which the interviewer feels may be
relevant. This is not possible in, for example,
questionnaires where the format of the
fact-finding exercise is pre-defined and fixed.
It is important, however, that the interview is
not allowed to digress too much or the
information collected may not be of
relevance. The one-to-one nature of interviews
means that the interviewer can clear up any
misunderstandings as they occur. It will
generally be obvious to the interviewer that his
question has not been fully understood, either
from a blank look or from an inappropriate
answer.
11
In an interview the researcher can easily
rephrase the question so as to elicit a more
meaningful answer. Disadvantages Practically
speaking, interviews are time-consuming and
expensive to organise. Depending on the
geographical spread of respondents, interviews
may involve a large degree of travelling which in
turn may make it impossible to do more than one
or two interviews per day. Questionnaires can
gather a large amount of information from a
widely located population, but are very
inflexible and require careful preparation to
avoid confusion and ambiguity If they are not
compulsory then the sample response may be
unrepresentative. With an interview, a systems
analyst can select whom he wishes to interview
and this could be important if we want to
validate the research he has conducted into
existing company documentation.
12
It is also an opportunity for the systems analyst
to get a response to the observations he has
made, in particular to any exceptions or
variation form the normal routine that he
noted. It is important when setting interviews
up that attention is paid to the environment in
which the interviews take place. It should be
quiet and relatively free of interruptions. A
shop floor worker may be reluctant to answer
questions openly if his line manager keeps
popping into the room! Also important is that
the interview is not allowed to go on for too
long - twenty minutes to one hour is sufficient.
Any longer than this suggests that the discussion
has digressed too much or that the conversation
has become bogged down. Interviews are a
powerful tool for gaining information from
others. By making people aware of the problem and
by making people feel that they are contributing
to the solution, the systems analyst can use
interviews as a very effective fact-finding tool.
13
QUESTIONNAIRES Specially prepared document
designed to obtain a specific body of
information. Preparing the questions in
written form renders the information gathering
process impersonal. Unlike the interview
method, questionnaires are a rigidly structured
means to obtain answers to preselected
inquiries. No opportunities for in-depth
questioning when new facts are uncovered.
14

• Because of the impersonal nature of the
  technique, users do not feel genuinely involved
  with the systems development process.
• The preparation of a good questionnaire requires
  adequate time be spent on
• selecting questions
• designing their format
• arranging the document for the ease of the
  respondents.
• A good questionnaire has four characteristics
• 1.Validity 2.Reliability
• 3. Face validity 4. Ease of use.

15

• 1 Validity
• Validity of a data collection method means that
  it obtains accurate information that fulfills the
  investigators needs.
• Judged by whether it asks the right questions
  from the right people.
• To authenticate the validity of a questionnaire,
  the document
• 1 undergoes a pilot test with only a sampling of
  potential respondents prior to its mass
  distribution.
• 2 Then revised to
• remove any ambiguities
• include additional questions if necessary.

16

• Validity(contd)
• Pilot test may be repeated to guarantee that the
  revised questionnaire meets the goals of the
  study.
• Depending on the complexity of the questionnaire,
  it may have to be refined and tested several
  times.
• After the questionnaire has been administered to
  the entire survey population, the results should
  be verified by other methods.
• Can be validated by
• interviews with a selected group of users
• observation of the existing system
• comparison of the results with managerial
  projections

17
2 Reliability Measured by the consistency of the
responses. To ensure that the responses were
correctly given, more than one question may be
used to elicit the same information. The
answers to these redundant questions are checked
to see if identical data were obtained. If
not, the information collected by the
questionnaire should be considered unreliable and
an unacceptable basis for any decision-making
regarding the new information system.
18
3 Face Validity Provides the appearance of
eliciting valid data to those viewing the data
collection procedure. Eg if the respondent
thinks that the questionnaire asks the right
questions in the right way Questions that seem
meaningless to the respondent may evoke answers
that are given without sufficient thought. If
poorly prepared, the entire questionnaire may be
dismissed as a waste of time and tossed aside or
answered carelessly. Again testing the
questionnaire with a small group of users should
eliminate such problems.
19
4 Ease of Use Easy for the respondent to
complete in a reasonable length of time The
information it supplies is easy to synthesize
into a comprehensive description of the existing
system. So questions should be clearly stated
and simple to answer. Use closed-ended
questions such as multiple choice or short answer
questions, so that the answers can be quickly
tabulated. Open-ended questions are usually
avoided because of the difficulties in accurately
summarizing the responses.
20
Ease of Use(contd) The number of questions is
limited so that excessive time is not required by
the respondent. The overall appearance of the
questionnaire also affects the users ability to
understand the questions fully and issue a
suitable response. If the questions are badly
organised, the user may experience difficulty in
comprehending them Questions should be grouped
by subject area to allow the user to follow the
questionnaires overall purpose. This aids the
user in determining the correct response and also
contributes to the face validity of the document.
21
Advantages of Questionnaires They are a popular
way to elicit information. Once the
questionnaire is validated, information from
large groups can be procured conveniently and
economically. When the questionnaire is
properly designed, the information supplied by
the respondents can be quickly tabulated into a
cohesive document. When used appropriately,
the questionnaire is a valuable tool.
22
Disadvantages of Questionnaires A good
questionnaires require the investment of large
amounts of time and effort Proper construction
of multiple choice or short answers is often
difficult. If a multiple choice question does
not offer a suitable choice to the respondent,
the final tabulation of responses will be flawed.
Badly worded questions can quickly mislead the
respondent and yield incorrect information.
23
Disadvantages of Questionnaires(contd) If
open-ended questions are used, the task of
interpreting the responses and tabulating the
answers is laborious. At best, questionnaires
supply only the information requested. The
opportunity to discover new information in areas
unfamiliar to the systems analyst is not
available. The advantages of flexibility and
adaptability associated with the interview method
are not attainable.
24
Observation Most time-consuming of all
fact-finding techniques Important that the
researcher has a clear purpose as to what
information he wants to find out or what
particular phenomenon he wants to observe.
Observation entails watching staff go about
their tasks and duties. In general, people do
not enjoy being watched, so this is one of the
analyst's more sensitive tasks. It is not
unusual for people to modify their behaviour,
either consciously or subconsciously when being
watched
25
For these reasons observation may not provide a
true picture of events. Used in conjunction
with other fact-finding techniques, however,
observation can be used to check that what people
say they do and what they actually do are one and
the same. Observation shows up things like
interruptions to the normal flow of work. When
asked to describe their work, people will
generally describe a typical day. They are
unlikely to mention to exceptional
circumstances.
26
When replacing one system with another, it is
essential that these exceptions can be dealt
with. Observation also highlights informal
communication channels how staff make use of
documents, files etc., the balance of the
workload on a system over a given period of time.
Obviously observation not appropriate if staff
are not to be disturbed.
27
Reading The problem with this fact-finding
method is that it is difficult to decide when to
stop reading! Most organisations have an almost
limitless amount of written material Not all
material will be directly relevant to the
investigation. Seeks the advice of those people
who presently work in the area as to what
material will prove to be most useful.
28
Find Preliminary planning and investigation of
Systemwas done before it was introduced. It is
useful to go back to these previous surveys and
investigations to learn about the system. With
older the material, however, findings need to be
verified very carefully . Eg Company
instructions structure charts job descriptions
1. will yield important information regarding the
organisation's intended operation.
29
2. may help in identifying where actual practices
have strayed from what was intended.
3. Obviously only appropriate if there is an
existing system.