Essay/Term paper: Methods of social investigation

Essay, term paper, research paper:  Humanities Essays

Methods of Social Investigation

Emma Woodmansee

'Describe how you would plan and undertake an investigation into why some of
this College's students do not complete their degree courses.' (You have been
given only 100 to finance the study; and one term's sabbatical.)

Define the variables in the given title

After a Research Statement has been formulated it is very important that the
researcher defines any variables within it. A variable is any word whose meaning
may be ambiguous or which could have several different meanings. This is a
crucial stage in the planning process as a vague title renders any results at
the end of the research without true meaning.

In this case, the Research Statement is the given title "Describe how you would
plan and undertake an investigation into why some of this College's students do
not complete their degree courses.' Within this Research Statement there are
several variables : "college's', "students', "complete' and "degree courses'.
These variables will be defined as follows:

`College's' We will take this to mean students at Royal Holloway
and Bedford New College, University of London.

`Students' Undergraduates on a first degree (excluding post
graduates and so on).

`Complete' Graduate

`Degree courses' The course for which the student originally registered.

By defining the variables above there can be no confusion as to the meaning of
the Research Statement. This process also helps the researcher to focus on the
group of people that he wishes to study.

Decide on the purpose of the research

Having defined the variables in the Research Statement, the researcher now needs
to focus his attention on the purpose of the research, and consequently lay down
the Research Objectives. This part of the planning process allows the researcher
time to consider what he hopes to achieve from the research and ensures that the
research represents his objectives.

The purpose of our research is to identify the reasons for students failing to
complete the degree course for which they were originally admitted (variables
already identified). The results of the research would allow the college to take
action to encourage students to continue their studies and could even be used to
aid the selection process and perhaps prevent problems from the outset. This is
the final purpose of the research.





Who is to be studied

The researcher needs to identify the group of people upon which to base the
study. The process is made easier by the fact that we have already defined the
variables in the Research Statement. The research group has been thus so far
defined as those students of Royal Holloway and Bedford New College, University
of London who did not graduate from the first degree course for which they
originally registered. To concentrate the study group further, the research will
be based on those students who left during the academic year 1995/6 only. The
study will include students who were registered in all faculties within the
University.

Initial sources of information

Once the research group has been identified the researcher needs to consider how
to identify those individuals which fall within the specified group. In this
case, the information we require will be held in the College Registry and also
in the Faculty Offices. This information is confidential and is not available to
the public, and this difficulty will be dealt with in the next section.

How to begin the investigation

This particular research project requires the cooperation of a group of ex
students of the college on what may be a sensitive subject. Therefore, the first
action that we would take is to send a contact letter out to all those within
our potential research group. The letter would describe the research that we are
carrying out, it's purpose and the method by which we intend to conduct it
(thereby informing the recipient of what they will be asked to do). It would
then ask the ex student whether they would mind taking part in the research. It
would then ask those who are willing to be involved to send some basic details
about age, sex, faculty and so on.

The letter would not be personalised as we do not yet know the names and
addresses of the ex students. We would then take the letter (and copies) to the
registry and ask if they would mind sending them to the ex students that they
have on their files. This is to protect the confidentiality of the information
held at the Registry. We would also include a stamped addressed envelope in the
letter to encourage the ex students to respond. At this time we still do not
know the names of the ex students and will only know that information from the
response. I would also set a final response date so that we have a cut off point
and know when we can begin the investigation in full.

After the final response date we would examine the basic details of the
respondents (age, sex and so on) and from this information, formulate a Random
Sample. Random sampling is a procedure in which bias is removed from the sample.
In other words, we ensure that we do not have thirty women and 3 men in our
sample. We would aim to have a wide cross section of people with no one
particular group being any more predominant than the other.

In official statistics approximately 3% of all students "drop-out' of University.
Whilst I suspect that the figure would be lower here, we shall use this figure
as a base and given the number of students is approximately 5,000 would estimate
the total number of "drop-out' cases to be around 150. Bearing in mind that the
sensitive nature of the research may cause a low response rate we would hope to
have approximately 50 research subjects after the random sample.

Which method of investigation to use?

"A survey is a method of collecting information from people about their ideas,
feelings, plans, beliefs, social, educational and financial background". The
survey would be the most effective way of conducting this investigation for
several reasons. Firstly, the postal survey allows us to cover a wide
geographical area within our limited budget. Secondly, the survey allows the
respondent to answer the questions in privacy, which is important given the
sensitive nature of the questions and thirdly the survey allows us to correlate
the results more easily.

Survey design

Survey design is a complicated topic as so many points need to be considered
together. The points that we would bear in mind whilst designing our survey are
as follows:

The position of the questions is very important

The first question should be clearly related to the purpose of the
research as defined in the research statement.
The questions should move from the familiar, such as name and so on, to
the more complex.
Sensitive questions should be placed in the middle, so that the
respondent is still concentrating on the questionnaire but is not put off from
starting to answer it.
The final questions should be relatively easy to answer as the
respondent may be getting bored with the questionnaire.
It is very important to ensure that the questions are independent from
each other.
We should be careful about the language that we use, avoiding emotional
language, leading or presuming questions.
It is also important to stress the confidentiality of the questionnaire
so that the respondents are encouraged to both take part and answer honestly.

Bearing in mind these design criteria we would use open ended questions to ask
the respondent for basic information about their age, sex, ethnic origin, course
title, and year of entry to the college. We would then go on to ask them why
they decided to leave the college prior to graduating.

Pilot survey

After we had designed the survey we would conduct a small pilot survey on
current students at the University. A pilot survey allows the researcher to
check that the questionnaire is understandable and so on. It may also give him
some idea as to the amount of time it may take to get a response and may even
indicate a response rate. On a limited budget, it is hoped that the pilot survey
would help to prevent any expensive mistakes later on in the research project.
If the pilot survey is successful we would go on to send the survey, with a
stamped addressed envelope, to our sample group.

Survey Results

After the final deadline for respondents has passed we would collect together
the received surveys and initially examine the basic information (age, sex and
so on) to ensure that we still have a random sample.We would then look at the
reasons given for leaving the college. The easiest way to correlate these
results would be to catogerise them in a table such as that displayed below.

Reason for leaving Number of respondents Total
Change of course IIIIIIIIIIIII 13
Change of college IIIIIII 7
Illness IIIIIIII 8
Not ready to study/year out IIIIII 6
Financial/grant problems IIIIIIIIIII 11
No longer wishes to study II 2
English language problems I 1
Accommodation II 2
Other I 1

After we had correlated the results in a table we would calculate percentages to
give a more meaningful picture. We would then represent the results graphically
in either a pie chart or a bar chart so that the information can be assimilated
easily by anyone interested in the results of the survey.



Problems with our approach

The main problem with research of this kind is the sensitive nature of the topic
and the low response rate that might result from that. Also, those people who
reply to the initial contact letter may differ from those who do not and a whole
strata of ex-students may not reply to the contact letter because of the reason
that they left. This must be taken into consideration when the results are
published. A final problem may be that a large number of students may have left
the address held by the college and this may affect the response rate.

Conclusion

In conclusion then, this method of planning and undertaking an investigation
into why some of this college's students do not complete their degree courses
should produce some useable and unbiased results.

BIBLIOGRAPHY

Fink, A & Kosecoff, J (1985) How to Conduct Surveys

Giddens, A (1993) Sociology

Howard, K & Sharp, J (1983) Management of a Student Research Project

Kane, E (1985) Doing Your Own Research

Moser and Kalton (1971) Survey Methods in Social Investigation

Oppenheim, A (1992 2ed) Research Methods in Social Relations