Writing Your Dissertation How to Plan, Prepare and Present Successful Work

8/14/2019 Writing Your Dissertation How to Plan, Prepare and Present Successful Work

1/145


2/145

Writing YourDissertation


3/145

I f you want to know how . . .Writing an AssignmentProven techniques from a chief exam iner that really gets results

Critical Thinking for StudentsLearn the skills of critical assessment and effective argumentHow to Pass Exams Every Time

Proven techniques for any exam that will boost your confidenceand guarantee successWriting an EssaySimple techniques to transform your coursework and examination

howtobooksPlease send for a free copy of the latest catalogue:

How To BooksSpring Hill House, Spring Hill Road, Begbroke,Oxford OX5 1RX, United Kingdomemail: [email protected]://www.howtobooks.co.uk
http://www.howtobooks.co.uk/http://www.howtobooks.co.uk/


4/145

Writing YourDissertationHow to plan, prepare andpresent successful work

3rd edition

DEREK SWETNAM

howtobooks


5/145

Published by How To Content,

A division of How To Books Ltd,

Spring Hill House, Spring Hill Road,

Begbroke, Oxford 0X5 1RX. United Kingdom.

Tel: (01865) 375794. Fax: (01865) 379162.

email: [email protected]

http://www.howtobooks.co.uk

All rights reserved. No part of this work may be reproduced or stored in an

information retrieval system (other than for purposes of review) without the expresspermission of the publisher in writing.

The right of Derek Swetnam to be identified as the author of this work has been

asserted by him in accordance with the Copyright, Designs and Patents Act 1988.

Copyright 2004 Derek Swetnam

First published in third edition paperback 2000Reprinted 2001 (twice)

Reprinted 2002 (twice)


Revised and reprinted 2004


Reprinted 2006 (three times)

Reprinted 2007

First published in electronic form 2007

ISBN: 978 1 84803 126 5

Cover design by Baseline Arts Ltd, Oxford, UK

Produced for How To Books by Deer Park Productions, Tavistock, Devon, UK

Typeset by Kestrel Data, Exeter, Devon, UK

NOTE: The material contained in this book is set out in good faith for general

guidance and no liability can be accepted for loss or expense incurred as a result of

relying in particular circumstances on statements made in the book. The laws and

regulations are complex and liable to change, and readers should check the current

position with the relevant authorities before making personal arrangements.


6/145

Contents

List of illustrations ixPreface to the third edition xi1 Before You Start 1

Checking the requirements 1Choosing a subject 5Checking feasibility 6W atching ethical practice 7Planning realistic time scales 8Producing your master plan 11Checklist 13Case studies 14

2 What is Involved? 15Judging how much work 15Planning your tasks 16Writing, revising and editing 18Devising a schedule 19Keeping a sense of proportion 21Assessing your own proposals 22Assessing reliability, valid ity, generalisability 23Using tutors and support 26Adopting house styles 27


7/145

W R I T I N G Y O U R D I S S E R T A T I O N

Checklist 28Case studies 293 Getting Started 31

Selecting your research methodology 31The researcher's status 41Sampling 42Following general patterns 46Structuring your dissertation 46Reading for a degree 48Checklist 54Case studies 54

4 Techniques 56The tool box 56M aking observations 57Preparing questionnaires 58Conducting interviews 64Ne gotiating with people 69Checklist 71Case studies 72

5 Down to Details 74The introduction: setting our your stall 74W riting your literature review 76Writing about research methods 81Presenting and analysing da ta 83Presenting conclusions and recommendations 9Compiling the bibliography 93Appendices 96Checklist 97Case studies 97

vi


8/145

C O N T E N T S

6 Writing Up 99Choosing the style 99Checking spelling, grammar and punctuation 102Keeping a coherent thread 108Presentation, printing and binding 108Checklist 113Case studies 113

7 The Final Stage 115Mark your own dissertation 115Preparing for a viva 118W hat happens now? 120Repairing a disaster 121Checklist 121Case studies 122

Fu rther reading 125Glossary 127Index 131

v i i


9/145

This page intentionally left blank


10/145

List of illustrations

1 A typical task plan 172 A sample dissertation proposal 203 A sample tally sheet 594 Sample covering letter 70


11/145


12/145

Pre faceto the Th i rd edi t ion

Almost every higher education course in the countryrequires a thesis, dissertation or research project of someconsiderable length in f u l l or partial f u l f i lm e n t of itsquali f icat ion. Such an undertaking of ten strikes terror intothe average student, who is typically assumed to be able towork alone, to understand all the principles of research andto have high-level reading and writing skills. Where thesehave been acquired is not always apparent.Some institutions of fe r excellent research methods coursesand impeccable support; others, sadly, leave students to sinkin a sea of panic which leads to high rates of non-completionor drop-out.

There are many top grade books on research methods butmost assume that the researcher is a professional with u n-l imi ted time and resources. This book is based on the realexperiences of students who do not match such a descriptionbu t w ho need ideas f o r planning and producing a good pieceof work without stress and total disruption of their personall i fe.


13/145


The basics are: asking a question, finding the answer andreporting the answer. It sounds easy: it can be easy andenjoyable but only for the organised and committed student.Only minor changes have been necessary to this edition.Student feedback suggests that the book fulfils its purpose ofproviding signposts to the most direct route through what,for many, remains a daunting task.Mature students especially will appreciate the pragmaticapproach which still makes no concessions to the quality andintegrity of the final product.The temptation to expand this book into a 'great tome' ofresearch methods has been resisted! There are quantitiesof advanced texts in every library and new styles of inter-active study, internet usage and co-operative publishing haveemerged.Despite this, hardly a week goes by without a report of someludicrous, contradictory research into an aspect of diet ormedicine carried out at enormous expense.

This book remains committed to the practical fundamentalsof useful inquiry undertaken with honesty and integrity byordinary students.

Derek Swetnam

xii


14/145

Befo re You Start

CHECKING THE REQUIREMENTSCongratulations! You have been accepted for a higher-levelcourse or have reached the stage where a taught coursemoves into the dissertation, project or extended study phase.There are over 5,000 postgraduate courses alone offered byuniversities and colleges in Great Britain and countless othercourses demanding at some stage a potentially dauntingindividual research study.

Research is inquiry into some aspect of the physical, naturalor social world. It must be systematic, critical, empirical andhave academic integrity. All students should assume thattheir work will be subject to public scrutiny and whateverapproach is used readers must be convinced that all therecommendations and conclusions are firmly grounded inmeticulous work.You would certainly not have been admitted to your courseif the staff did not think that you could pass. The greatestdanger is not failure but non-submission and some courseswhich appear to have high failure rates in fact reflect thelatter.

1

1


15/145


This book is about planning, strategy and keeping yoursanity while producing good work.Although every educational organisation has its own idio-syncratic requirements they have many features in commonand general principles apply to all. Apart from the very bestPhD work all student research is a compromise betweenwhat is desirable and what is possible. Mature, part-timestudents will need to keep a fine balance between thedemands of job, family and course.

Mature students have special problems which they are oftenvery reluctant to share. They have a high ego stake in thequalification, especially if they are already in a senior post.There may be other more junior members of staff who havealready achieved the qualification. They often feel that theiracademic writing skills have atrophied and that they will beoverwhelmed by the phenomenal brain power of the otherstudents. Almost by definition the older student is moreprone to marital, family and health problems.Few of these fears are ever realised and are usually morethan compensated for by the wisdom and motivation thatcomes with age. Whatever problems you may have it iscertain that someone else has experienced and overcomethem. Beware of rationalising difficulties: that is blaminghome, job, health and family unfairly for your lack ofprogress. Confide in your supervisor and most will moveheaven and earth to assist.

2


16/145

B E F O R E Y O U S T A R T

Assessing your abilitiesIf your course is optional and especially part-time you willalmost certainly have terrible butterflies in the stomachabout your ability to see it through. This is absolutelynormal. Try the following self-assessment:1. Do you have the motivation and commitment?2. Can you cope w ith the dem ands of you r job and family as

well?3. Do you have the reading and writing skills?4. Can you face the travelling and possible evening work?5. Can you afford the course fees or obtain funding?6. Can you work on your own?7. Do you respond to pressure and deadlines?If that all seems a bit gloomy try this:

1. Are you looking for an exciting and rewardingexperience?

2. Do you want to expand your thinking powers?3. Do you want to meet stimulating people?4. Are you ready to enhance your career opportunities?5. Do you want to feel proud of your achievement?Convinced? Here is a tip from one of my students who hadevery possible personal problem during her research. Do notassociate your dissertation too closely with your work orprofessional life. Regard it as an escape from stress, not as anaddition to it; it really helps.

3


17/145

WRITING Y O U R D I S S E R T A T I O N

Building your confidenceA great preventer of progress is lack of confidence andthe insecurity that results. To start your confidence-buildingprogramme check all the details of the requirements andregulations as follows:

W hat is the exact date of submission?W hat wordage is required?Are there any intermediate specified dates, eg literaturereview completed by?How m ust the work be presented?W hat are the rules about form at?What tutorial support is available?

It is quite helpful to see actual examples of previous studentwork, available from the department or library, if only toconvince you that you can do better than m any of those thatwere accepted!It is strongly advised that the house style and format isadopted from the outset (see Chapter 6). Even if the studentis a computer expert, re-formatting many pages of text,especially if desperate for time, can be a frustrating task.The wordage in research projects varies from about 10,000up to 60,000 or more for a high-level pure research degree,the average at master's level being 25,000. There has beena distinct tendency in recent years for these figures tobe reduced. If the numbers produce a sick feeling in the

4


18/145


stomach be reassured that exceeding the limit is a far morecommon problem than the reverse.The crucial issue of using tutors will be returned to later.

CHOOS ING A SUBJECTThis apparently simple process can be agony. Usually onlygeneral parameters are given, yet great precision is expected.You will be seriously handicapped unless the topic is one inwhich you are interested enough to spend up to a year ormore and to sustain that interest during hours of work. Thedissertation should be 'about something': the days whenrambling, discursive essays were accepted have long goneand you must never start without a detailed plan, preferablyalso a title, even though this may be subsequently modified.

The selection processThe process is rather like sharpening a pencil, fo r example:General area of study - sociologyParticular interest - groups of old peopleMore specifically - community careEspecially - in residential homesPrecisely - in warden-controlled homesDraft title - 'The management of community care inwarden-controlled residential homes'.

The eventual fine point is the working title which may befurther refined or limited to one exact location. From this

5


19/145


stage another useful technique is to develop an ideas listwhich covers about 30 possibilities for the research, to beedited down as the research questions are developed. Somestudents do a personal brainstorm or draw multileggedspiders w ith an idea on each leg. If you have access to otherson the course that is a useful stage at which to exchangeviews w ith them and/or arrange an initial tu torial meeting.

CHECKING FEASIBILITYThe title and the ideas may sound wonderful but are theypossible? However worthy it may be to survey 50 localauthorities, can you support the postage, cost and travelthat such a project implies? One undergraduate proposedresearching the roles of Thomas Hardy and Arnold Bennettas regional novelists in six months: at least 15 novels to beread!It is more important for you to achieve the qualification andstay sane than to impress the academic world with an amaz-ing work of erudition. If appropriate, high-level scholarshipmay be developed later from the w ork in hand.Many of the ideas on the first list will have to be crossed outon the grounds of impracticability. Check the following:

Is the project physically possible in time, distance andvolume of work?Can it be afforded?

6


20/145


Do you have, or can you develop rapidly, the skillsneeded for the research?Will you be permitted to gain access to the suggestedsites?Can you find the necessary literature?Are there any ethical or moral problems?Will the topic remain 'live' over the period of theresearch?Would you have the support of the college and/or youremployer?

Do not assume anything. If in doubt about access get writtenconfirmation. A burning issue in schools or the healthservice can be extinguished by legislation. Many educationstudents have had their research made virtually worthlessby government legislation on such matters as school ex-aminations and governorship.

Be particularly careful if you plan to research with childrenor ethnic minorities. While most employers are helpful andcooperative some can be jealous or obstructive, especially ifth e work m ay reveal defects in an existing system.

WATCHING ETHICAL PRACTICEStudents may be surprised when they are given a writtenpolicy on the ethics of research and even asked to sign thatthey have read it. Some of these can be quite complex,especially where physical research on people or animals is

7


21/145


involved, but all students should follow some generalprinciples which are common to all policies. You have a dutyto ensure that:

no harm should come to participants in the researcheither physically, mentally or sociallyparticular care is taken not to exploit the vulnerability ofchildren, the elderly, the disabled or those disadvantagedin any wayno physical or environmental damage should be causedwherever possible participants are informed of the natureof the work and give their consentthe research follows equal opportunities principlesanonymity and privacy, where requested, are guaranteedand honourednothing is done that brings your institution into dis-repute.

PLANNING REALISTIC TIME SCALESRemember that you do not really have all the time thatyour institution has apparently allocated for submission. Onemust think positively but here are some factors that regularlycause the operation of 'Murphy's Law' which states that ifanything can go wrong, it will!

family problemsillness

8


22/145


holidays (yes, they are needed)computer failurescontacting tutorsrejection of manuscriptdelays in typing and binding.

If none of these occurs you have a bonus!Don't be pessimistic but be realistic. Students who areunable to use a word processor may find difficulties inemploying a typist (50 pence a page or more) and threeweeks may be required. Modern technology has made any-thing that is not computer generated virtually unacceptableand some institutions state this as a policy. Whatever yourlevel of competence it will be required for you to storefloppy disks, CDs, or DVDs; at this early stage it isabsolutely necessary to prepare back-up copies of everythingthat is produced. There must be two copies kept securely intwo different places; perhaps home and work. Professionalbinding can take from two days to two weeks.This may sound somewhat paranoid but the author haspersonal experience of students' irreplaceable work missingboth through theft and fire.A dissertation which is to be completed 'in a year' mustactually be done in nine months and most students find thatprocrastination can easily become a way of life. A generalguide for any extended project or research is to plan to use

9


23/145

WRIT ING YOUR DISS ERTA T ION

only 75% of the supposed time available, which should betentatively allocated as follows:Introduction 5%Literature review 35 %Research methods 10%Data collection 20%Analysis 15%Conclusions and recommendations 10%Bibliography and appendices 5%

This general balance holds good for a wide range of researchtypes. However, scientific and experimental work mayhave the data collection part extended, particularly in higherdegrees. The boundaries are not discrete and the readingand updating of literature runs throughout.

During the last few years there have developed differentmethods of submission of dissertations which at first seemquite revolutionary. These include more active styles ofwork involving colleagues or co-operative groups; jointpreparation and submission; 'critical friends' and mentors.Several universities now offer higher degrees by publication,where the student prepares papers for publication onthe research or theme. The papers may be spread over theduration of the course and are generally of a very highacademic order. Acceptance of a paper may depend onreferees who scrutinise the quality of the contribution. Theseless traditional modes of study are commonly found in main-land Europe: the universities may publish and produce thecompleted work.

1 0


24/145


If your institution has introduced or intends to introducesuch methods as the above, do not panic! They are not'easier' or 'harder' - just different. The fund am entals remainthe same: to contribu te significan t and original research forthe extension of human knowledge, with rigour and in-tegrity.It is gratifying to see that several of the new er universitieshave completed excellent guides to their courses that usenon-traditional data collection and assessment methods. Inone case over fifty precise guidelines are prescribed. Othersstill operate the 'sink or swim' system.

PRODUCING YO UR MASTER PLANThe choice of general area of study is easy as it is usuallydefined for us by circumstances, our abilities and the coursethat has been undertaken. It is the refinement, focusing anddevelopment of the research questions that present a hurdle.Take a deep breath and answer these three questions:

W hat do you want to know?How are you going to find out?W hat will you do with the answers?

From an amorphous, vague idea must emerge a sharp,precise plan. The initial notion will be modified by contactwith colleagues, reading and practical considerations whichwill provide you with a location in your own work area oroutside it. The purpose of the work should become clear: is

11


25/145


it to develop theory, to monitor practice, to evaluate, toincrease understanding, to recomm end policy?From this process are produced several research questionsand possibly a hypothesis. The latter is a premise basedon observation or known facts that provides a basis forempirical testing to prove or disprove it. While this may beessential in pure scientific research it is often inappropriatein education or social sciences where hypotheses can emergeas the work proceeds. However, there is no escape in anyarea from defining the problem or questions.You may be directed to propose a hypothesis. It is usuallycentral to experimental research although many greatscientific discoveries have been made without one! In thesocial sciences an over punctilious insistence on hypothesesoften leads to absurdity or verbal gymnastics to explain whythe original hypothesis was completely superseded. Basedon the definition above which of the following are validhypotheses?1. Hybrid animals within a species live longer than pure

bred ones.2. Boys who play the flute and girls who play the drums are

more likely to be bullied at school than children who takeup instruments considered 'more appropriate' to theirgender.

3. Prisoners should not be kept more than two to a cell.4. Intrapa rtum asphyxia is a major cause of cerebral palsy.5. St Lucy's High School needs a marketing policy.6. Open-cast mineral extraction has detrimental social and

economic effects on the immediately surrounding area.

12


26/145


A re there any that m ight have emerged from research?Answering the question, 'How are you going to find out?'demands some knowledge of basic research styles. Positivistwork is concerned only with observable, objective facts:interpretative or subjective work uses explanation andinterpretation. Associated terms are quantitative: numbersand measuring, and qualitative: qualities, description andappearance. Although some researchers take an inflexiblestance on these, in practice m uch valuable wo rk has areas ofoverlap. Note at this stage that a study of research methodswill be required as well as the general reading.This may all sound very onerous and so far not a word hasbeen written! It is nothing to the tangle that will result ifplanning is skimped. The master plan will show wh at is to bedone, why? where? and in wh at time?

CHECKLISTHave you:

checked and understood the general requirements?chosen, refined and focused your subject?examined its feasibility?selected a tenta tive title?suggested a time scale?proposed a location?discussed ideas w ith peers and tutors?

13


27/145


CASE STUDIESPaul chooses a subject close to hom ePaul is in his final year of a BA Geography course atuniversity. His prime interest is in hum an geography and hisresearch project of 10,000 words must be done in the finalyear while still completing taught elements of the course.The university allows only three blank weeks for datacollection. His initial suggestion of 'a look at out-of-townshopping' seems hopelessly difficult and likely to needextensive travel, but a chance remark by a colleague remindshim of the furore caused in his home town by open-castmining: a familiar area, some knowledge - that's it! A brain-storm produces ten ideas and a possible title: 'A study of thesocial and economic effects of open-cast mineral extraction'.Alison's topic is work-relatedAlison is the deputy head of a Roman Catholic, inner-citysecondary school. Despite the heavy demands of the job andhaving two small children she has completed two part-timetaught years of an MA in Educational Studies. A 20,000-word dissertation must now be completed in one academicyear as she was unable to produce anything last yearowing to fam ily pressures. The general parameters given are:curriculum, special education or m an age m en t. She is advisedby a tutor to 'focus on something accessible' and decides tomake an ideas list on the management of her o wn school.

14


28/145

What is Involved?

JUDGING HOW MUCH WORKIt is possible for an able student to write 2,000 words in amorning, but most such attempts end in disaster. All initialwriting should be considered as draft and the student mustnot develop an over fondness for any part. Whole areasmay have to be revised or ruthlessly edited out as thework develops. Although the dissertation is original it has tocontain many references to the work of others, quotationsand perhaps diagrams or tables. One difficult page may takean hour.You should try to do some work every week; even a fewminutes is important as it helps to m aintain motivation and asense of overall purpose and shape. If this is lost the workcan easily degenerate into a series of disjointed essays.Some students find that a disciplined time-table is useful,others find that too restrictive, but wh ichever style you ad op tregular production is essential.Depending on the level, you have between ten and twentymajor aspects to address including the focusing that hasalready been done. The research questions can be furtherrefined and should be constantly referr ed to before each new

15

2


29/145


area of work. Reading and recording the literature andresearch m ethods literature may be a third of the task.

PLANNING YOUR TASKSIt is not intended that the task plan, Figure 1, be slavishlyadhered to - individual circumstances vary. For example,some institutions strictly limit the students' choices. How-ever, all successful students follow the basic strategies.See if you can match some of the stages of the flow chart inFigure 1 with the following narrative:

A nursing student has read three general texts aboutthe relative effects of different types of care on newlyborn babies. The texts indicate that this area has beenresearched extensively by both amateurs and pro-fessionals but still interests him greatly.The student, in discussion with a tutor, focuses theproposed work more narrowly on premature baby care inhome and hospital. The research is to evaluate the presentsystem and to discover the differences between homeand hospital-based care, making recommendations.

The study will be based on Ward 27, MightownGeneral Hospital in association with the two seniornurses responsible.

The work will be a case study of the Post Natal groupof Ward 27 and careful negotiation will be required toobtain admission to sensitive areas. Close attachmentto one nurse tutor is suggested.

Circumstances dictate an opportunity sample of upto five mothers and babies in each type of care, and

16


30/145

WHAT IS INVOLVED?

Preliminary reading and choosing the study area

Establishing the focus, specific objectives and researchquestions

Identifying the place, material, institution, people tobe studied

Selecting the style of research

Organising and obtaining access

Defining th e sample

Choosing and preparing research techniques andinstruments

Closing the investigation

Sorting and studying the data

Interpreting and analysing the data

Transforming draft writing into finished product

Submission, collecting your qualification

Fig. 1. A typical task plan.

17

Collecting data


31/145

WRIT ING Y O U R D I S S E R T A T I O N

observation and interviews will be used to collect dataover a six-week period defined by the department.As well as helping with planning, the task plan mayserve vital functions such as clarifying thinking, refining andfocusing, and judging the balance between desirability andfeasibility mentioned in Chapter 1.Despite the apparently restricted scope of the above study,some fairly high-level skills and knowledge will be required.Even at stage 2 warning bells may be ringing about suchissues as access, ethics and academic levels. It is better tomake a short trip back to the drawing board at stage 1 or 2than a stressful journey after several weeks.How far do Paul and Alison (the characters in our CaseStudies) follow the Figure 1 Task Plan?

WRITING, REVISING AND EDITINGA question often asked is w he the r the dissertation should bewritten up as the research proceeds or at the end. Whilethere are undo ubtedly people who can m ake sense of piles ofnotes and cope with hours of writing, most cannot. It is alsovery reassuring to see whole completed sections, subjectalways to revisions and amendments as the work develops.Yo u are strongly advised to write up as soon as the materialis ready. There will be many occasions when you will re-readeach chapter or come across new and valuable additions toyour stock. Unorganised material gets into an even greatertangle when tinkered with, whereas a good typed copy can

18


32/145

WHAT IS INVOLVED?

be marked with a pen and easily seen. Resist the temptationto edit at odd times: keep an edit list with the page numberof each edit, omission or error and tackle each chaptersystematically.

DEVISING A SCHEDULEYou may be required by your institution to provide aschedule or proposal either devised by yourself or accordingto their format. In either case and to assist your o wn think-ing, the following outline should be useful:

What are the theme, purpose and main objectives of theresearch?Where will it take place?What is the hypothesis (for scientific research)?What is the theoretical basis of the research? (You mayneed tutorial advice fo r this.)What general, preliminary reading has given yo u ideas?What might the research lead to o r be useful for?

At higher levels, such as MPhil o r PhD, greater complexitywill be required and you may have to present your sub-mission to a committee.The great enemy here is vagueness. Avoid words like 'havea look at', 'try to' and 'see if; be positive and examine,measure, evaluate, survey and assess.For an example see Figure 2.

19


33/145

DISSERTATION PRO POSA LDraft title

An evaluation of community facilities at Hightown LeisureCentre: their use, management and development.Aim of the researchFollowing recent criticisms from organisations and the public

that the use of the new leisure centre is restricted to certaingroups, the study aims to investigate the present use ofthe centre and to determine whether some people are dis-advantaged by the facilities currently offered.

The study will concentrate particularly on the use made ofthe centre by mothers of small children, ethnic minoritygroups and the elderly over a six-month period. Thehypothesis is that greater community use would result if moreattention was paid to the needs of these groups. It is intendedthat the research will result in practical management recom-mendations.

Questions to be addressed include:1. Are there perceived and real inadequacies in provision?2. Do they disadvantage certain social groups?3. Does this have an undesirable effect on communityrelations?4. How may provision be made more equitable?

Style and techniquesThe work will be a case study generalisable to other similarcentres. Use will also be made of survey and documentaryanalysis. Instruments for data collection will principally beinterview and questionnaire to obtain descriptive statistics.

Theoretical base and initial readingReference will be made to management theory, managementof change and multicultural education.

Field (1991) Post 16, community education and racialequality. Multicultural Teaching

Symonds and Kelly eds (1998) The Social Construction ofCommunity Care

Haworth (1997) Work, Leisure and Well-beingFig. 2. A sample dissertation proposal.


34/145

WHAT IS I N V O L V E D ?

Other headings that you may come across include:anticipated problems, equipment, ethical dimensions andcosting.

KEEPING A SENSE OF PROPORT IONReference has already been made to checking the preciserequirements for the project. It is also essential to knowthe proportion of marks that are awarded to the dissertationwithin the total qualification. Variations may be between10% and 100%. Even in a taught master's degree thethesis mark may be only one-third of the total; however,inevitably, passing this section is a condition of obtaining thedegree.This knowledge should help you to allocate time tocomplete the thesis without losing your friends, family andsanity! It is not a great idea to end up with both yourqualification and a divorce.It is important to get the co-operation of those near to you atcollege or at hom e so that if you allocate a par ticula r eveningor S aturd ay morning to the project, it is inviolable.Experience shows that most successful students distributetheir workload in the following pa ttern over 12 mo nths (prorata for shorter times):

Reading, note-making, planning, setting up systems,writing the introduction: three months.Writing the lite rature review: two months.

21


35/145


Refining and writing up the research methodology: onemonth.Carrying out an d recording em pirical w ork: one m onth.Analysing data: one month.Preparing conclusions and recommendations, appendices,bibliography, final amendments: one month.Proofreading, corrections, binding: one month.

ASSESSING YOUR OWN PROPOSALSBe able at any time to answer the questions 'What is thisresearch for? What is its value? W hat will it add to theory orpractice in the real world?' If you are unable to do so,the focusing suggested in Chapter 1 has not been carriedout properly and this will be reflected in vague researchquestions and a 'woolly' title. Here are some examples ofrejected (a) and subsequently refine d (b) titles:

l(a). A survey of appraisal in industry(b). The management and improvement of appraisalsystems in the Hightown Division of BerylmagPharmaceuticals

2(a). Equine veterinary medicine(b). A study of the changes in veterinary medicine

resulting from increasing recreational horse riding

Y ou will find that the remaining two months mysteriouslydisappear.

22


36/145


3(a). Tem perature m easu rem ent in the steel industry(b). A study of the accuracy and durability of refractoryshielded thermocouples in blast furnac es

4(a). The popularity of small schools(b). Marketing a small school: educational and social

advantages as perceived by parents and stake-holders

You might like to consider whether the above need furtherpolishing perhaps by the addition of a sub-title.

ASSESSING RELIABILITY, VALIDITY, GENERALISABILITYThese three concepts help provide blunt answers to thequestions 'Is it any good; is it any use? ' They are not theexclusive province of the positivist, scientific, quantitativeresearcher. The qualitative worker has to compete withwhat has been called 'The lure of numbers ' - that is, theunjustified belief that data involving measurement areinherently more valuable than things that are observed ordescribed. This prejudice makes it even more important forthe qualitative researcher to subject all data to rigorousexamination:

Reliability: would the same procedures, experiments oractions carried out again produ ce the same result?Validity: are we actually measu ring or observing w hat weclaim to be?Generalisability: is our wo rk applicable or useful to otherpeople or situations?

23


37/145


The physical scientist may be required to provide complexmathematical or statistical analyses of the above, but thework of the social or educational researcher is alsoproblematic. Compare the task of finding out whether onegramme of silver nitrate always yields the same amountof precipitate when treated with sodium chloride solutionwith that of discovering what percentage of residents in avillage believe in God.All residents? What do you mean by'believe'? Which God? Do they tell the truth? If reliability ischecked by repeating the research in a year's time, is thesample the same although it is a year older?These problems are not insuperable but must be acknow-ledged and considered to maintain academic integrity.Beware of variablesWhenever human beings are involved it is very easy toproduce apparently convincing research that is quite invalidbecause the variables are incorrectly identified. Suppose itis observed that people with tattoos are less successful inexaminations than those without. A n unwise researchermight spend time investigating the nature and physicaleffects of tattoos, whereas the true variables are certainlya complex mixture of social, psychological and attitudinalfactors. The research w ould have no validity.Try listing some factors that might influence the reliabilityand validity of research into the effects of alcohol on drivingperformance. You should have no difficulty in finding ten.A critical approach to research can often uncover somesurprising variables. If you were researching the occurrenceof domestic house fires would you collect data on the

24


38/145

WHAT IS INVOLVED?

presence of inflammable furniture, open fires and air flowsin houses? Research by the Cheshire Fire Brigade, in 1992,showed that a major variable was the degree of family stress.In pure experimental research you must clearly distinguishbetween the independent variables (those manipulated bythe researcher) and the dependent variables (those whichchange as a result of the manipulation).Generalisability, or generality in physical science, is particu-larly important to the social scientist. How far might theresults of this work offer guidance or suggestions of practiceto others in similar situations? Even such an apparentlyrestricted study as title Ib above may enable readers toextrapolate to their own situation.

You may feel that your study is too brief or at too low a levelto need attention to these concepts, but some understandingwill pay dividends in assisting clarity of thought and purpose.A serious danger faced by social and physical scientists is'Knowing the answer before you start,' or ' OK, here's theanswer, now what's the question?'. This principally occurswhen a central issue is almost universally perceived as bad ordestructive. Did you find ten factors involved in researchingthe effects of alcohol on driving? Try some of these on yourfriends:

How can we stop the rising number of child casualties onthe roads?( in reality, the number of child deaths on the road hasfallen by 23% in fifty years)

25


39/145

WRIT ING Y O U R D I S S E R T A T I O N

Punishing the speeding motorist should be our priority inreducing accidents( fewer than 10% of accidents can be attributed to speed)It is better to integrate handicapped children into main-stream schools(better for whom?)Equal Opportunities laws have stopped discriminationagainst women(Women occupy only 7% of executive posts in theprivate sector)Examinations should be replaced by course-work assess-m ent to im prove the chances of w orkin g class pupils(there is evidence that course-work assessment favoursmiddle-class students)

Yes, of course these examples are simplistic but such flawedthinking is seen frequently in students' work. A particularcourage is required by a researcher when an organisation hasalready spent millions on the 'wrong answer ' .

U S I N G T U T O R S A N D S U P P O R TPractice varies widely among and within institutions. Youmay find one or more of the following tutorial practices:

unspecified time on demanda stated time allocationsome com pulsory g roup or individual tutorials

26


40/145

WHAT IS INVOLVED?

shared tutorial responsibilityan initial tutorial only.

Find out the rules and, unless you are supremely confident,use the time. Experience over many years shows that failuresand non-submissions are heavily dominated by those whohave declined to see tutors. In addition, a top secret piece ofadvice: if you fail and appeal, the fact that you have notaccepted tutorials will count against you. Most tutors take afierce professional pride in their students and make them-selves easily available; others may be more elusive.It is vital to prepare for each tutorial by listing the areas thatneed discussion and the precise points that should be clearedup. If advice is sought about a particular chapter or sectionmake sure that the tutor has been provided with a draft copyin advance. Do not make tape recordings of the sessionunless the tutor agrees.Try to strike a friendly but strictly professional relationship.Some tutors refuse to make technical corrections to workas a matter of policy. You may have a preference whichyou can make known. Always try to space tutorials andamounts of script submitted; it is unfair to expect staffto deal effectively with thousands of words at short or non-existent notice.

ADOPTING HOUSE STYLESAs well as checking requirements, right from the start adoptthe layout and format recommended by the institution, even

27


41/145


in dra f t f o r m . More stress is caused by wasting t ime o ntechnical al tera t ions tha n by m akin g academ ic revisions.In the absence of guidan ce, use the g eneral pattern discussedin Chapter 3 and the conventions as follows:

paper: white A4double space all main text, single space indentedquotat ionsfonts: plain and s t a n d a r dmargins: 1.5 inches to the left, 1 inch elsewhere.

CHECKLISTHave you:

est im ated the wo rkload ?clarified yo ur research questions?wri t ten out a pro ject schedule?unders tood what is specifically requ ired ?critically assessed your own proposals?considered the likely va lue and use of the w o r k?m ad e con tac t w i th your tu to r?thought about your family, f r iends and personal life?

28


42/145


C A S E S TUDI E SPaul is encouraged to be realisticOn checking the university department's regulations Pauldiscovers that he is only allowed one tutorial for his projectso he starts to organise a written submission. He quicklyrealises that his great idea is hopelessly over-ambitious andwould probably tax a PhD student. This is confirmed by thetutor and together they settle on one research questionfocused on one site only: W ha t are the immediate effects ofsmall-scale mineral extraction on a small community fromthe social and environmental aspects?On a weekend visit home he decides to visit the site but istold politely but firmly, 'No admittance without an officialuniversity letter of introduction'.Alison focuses on marketingAlison has decided that the central management problemfacing her school over the next five years is that of marketinga denominational school in a highly competitive situation.Her family agrees tha t Saturday m ornings, at least, will be asacred writing day. The title suggested on her submission is'Improving the marketing of a denominational high school:management problems and strategies'. The headteacherreceives the idea enthusiastically - perhaps too enthusiasti-cally - and her tutor accepts the focus. Several researchquestions emerge, the answers to which should havepractical value to the school:1. What are the perceptions of parents and stakeholders

about educational standards at St Lucy's High School?

29


43/145


2. What factors are influencing parents to choose or rejectthe school?3. What special problems occur in marketing denomina-

tional education?4. How could management strategies for marketing be im-

proved?

30


44/145

Getting StartedSELECTING YO UR RESEAR CH METHODOLOGYThe basic question tackled here is 'How am I going to f indout the answers?' The issue must be addressed by everystudent irrespective of the degree of formality required bythe institution. At higher levels a complete justification ofthe approach is required including a survey of possibilities, arationale for choice and an extensive review of the chosenmethod and its relation to the variables; a minimum require-ment is an explanation of the reasons for the approach an dits likely advantages an d disadvantages.The literature can be confusing in its terminology, especiallyin the use of the terms 'methodology' and 'method', some-times interchangeably. Distinctions may also be foundbetween 'strategies' and 'procedures'. You need to cover twoareas: the overall style of the research and the actual tech-niques to be used within the empirical work.

Research falls into tw o basic styles which are objective andsubjective (or nomothetic and idiographic). Objectiveapproaches are concerned with physical characteristics andthe external world, universally applicable rules and laws,tested through hypothesis, experiment and survey.Subjective approaches deal with the created social lives of

31

3


45/145


groups and individuals through observation and explanation:both are systematically controlled and empirical and may beused by physical or social scientists.A term that students may find puzzling is phenomenology,which is an extreme subjectivist, qualitative mode of inquiry.The phenomenologist abandons all prior assumptions aboutthe social world and does not test any hypothesis asthis would automatically imply a preconception about theactors and the situation. This perspective demands a stylewhich recognises the distortion due to the researcher inevery situation, There are many levels o f perceived reality inhuman experience which can be studied using subjectivistmethods and techniques. A phenomenologist would notemploy techniques such as structured interviews.It is simply not true that the amateur researcher hasthe same freedom of choice as the professional. There areconstraints of time, cost, access and ability amongst othersbut these need not demean the work providing academicintegrity is maintained by acknowledgement of thelimitations. More than one methodology may be used withareas of overlap.The institution may specify a research method, for examplethe experimental method in pure science projects, or alimited range of methods. It is quite common in nursingand midwifery degrees to limit the student to experimental,action or evaluative approaches.The student is not expected to write at length about thestyles that have not been selected but to concentrate on

32


46/145

G E T T I N G S T A R T E D

the one that has. A brief outline of some possibilitiesfollows,Action research: sometimes called 'participative'In this style the researcher tackles a real problem, inter-venes, makes changes and monitors the effects. The subjectsparticipate and implement the interventions, leading tofurther changes in a spiral pattern.An example might be a hospital ward sister and staffinvestigating the effects of organisational changes on patientwelfare and recovery rates over a three-year period, makingchanges to systems at regular intervals and modifying aftereach intervention. The sheer practical difficulties of such aproject are immediately apparent.General criticisms of action research include the lackof detachment of the participants, weak identification ofvariables, the time scales involved and the chance that thework m ay degenerate into haphazard tinkering. However,meticulous action research can product satisfying, 'real life'problem-solving.SurveysSurveys attempt to gather information from an entire group,or more usually a sample, which can then be used to makeinferences or generate policy or reveal unsuspected facts.Students often grab at this one, as superficially the processseems easy. In fact, as in all research involving people, it canbe a minefield. Even the last National Census mysteriouslylost several million citizens!

33


47/145

34/ WRIT ING YOUR DIS SER TATI ON

The information may be gathered in several ways: forexample, interviews, which are face-to-face exchanges withparticipants, and questionnaires, or structured lists of writteninquiries, but it is easily invalidated by poor sampling andambiguities (see Chapter 4). Students are often disappointedby very low response rates to their surveys but should realisethat respondents may have little motivation to reply or maysimply be irritated by the whole process if it is difficult tounderstand or too complex. Subjects may lie or try to pleasean interviewer and causality can never be proved: '20% ofschool pupils dislike school because of school uniform'.Difficulties are neatly illustrated by two genuine quotationsfrom published research reports:

'Although more than 3,500 questionnaires weredelivered to tenants, who were offered the chance ofwinning a prize for taking part in the survey, only 373completed questionnaires were returned.''The agents throughout North Staffordshire refused arequest by health experts to send questionnaires tocustomers asking them if they had followed the adviceon safe sunbathing.'

Surveys can be purely descriptive: how many males over 21were unemployed at the end of 200x? or explanatory: whatwere the claimed reasons for unemployment? They canbe purely for the collection of factual information or fordecision-making. In the latter case there are some fairlycomplex mathematics used for determining sample size,measurement error and analysis of data. If you are


48/145

GETTING STARTED

convinced that your study requires such treatment, specialistreading and tutorial help will be needed.Remember that all data collection has a cost in money ortime and that quite large extra costs may only result in minorimprovements in the quality of the data obtained. Bluntly,only do w hat you have to do.There may also be ethical factors: should the subjects be toldof your intentions? should they be offered anonymity? howdo you treat potentially sensitive biographical information?A great temptation to be avoided is the collection of greatamounts of data that have little, if any, use. Do you reallyneed dates of birth or marital status?If the survey method is chosen you must critically analysethe proposals:

What hypothesis is being tested or developed?How does the literature illum inate this?How could th e study be designed to provide the bestdata?What instrum ents are ap prop riate?How will the survey be checked and piloted?Is the sample representative?W hat is going to be done with the answers?

Poor quality surveys often represent the worst of research. A1994 'professional' survey concluded that the people of a

35


49/145


northern industrial city were aimless and apathetic, based ona sample of 10!Experimental methodExperimental method is used in all types of research but inits 'pure' form is associated principally with work in thephysical sciences. Some argue that all studies outside therange of physical science are more accurately described asquasi-experiments.

The process, in outline, is the formulation of an hypothesisbased on observation and theory that can be tested andproved or refuted. Independent variables are preciselymanipulated to prove causal effects on dependent variables.The whole is characterised by precision, accurate measure-ment and careful duplication which should enable exactprediction and generality. Such purity is rarely found inpractice and many great scientific discoveries have beenmade by accident or serendipity.

Experiment in the social sciences faces difficult issues ofcontrol, ethics and external validity. Empirical knowledgeabout people - numbers, ages, heights etc - is oftenwhat interests us least, whereas beliefs and attitudes defymeasurement.

Suppose an experiment is set up to test the hypothesis thatchildren's reading development is proportional to numbersof books available in classrooms. Apart from the difficultiesof ensuring equal starting points, equal motivation and equalability there is the insuperable ethical problem of setting upa control group which is completely deprived of books!

36


50/145


Nevertheless, there have been many fascinating and usefulexperiments with people, especially in the field of psychol-ogy. Even in this discipline where tight control is exercised,samples mature or are 'lost', practice changes results and thewell-known Hawthorne Effect occurs where the behaviourof subjects alters w hen they kn ow that they are taking part inan experiment.Ethnographic researchThis is a qualitative, descriptive style in which a researcherattempts a study of the behaviour and customs of a groupof people by integration into the society. The method hasproduced many riveting studies but is unlikely to berecom m ended to ma ny stu dents as i t is t ime consum ing andthe procedures are actually very complex.Criticisms often made of the value of ethnographic workinclude: it is unsc ientific, it is not generalisable, thepersonality of the researcher causes bias, small groups areunrepresentative. Proponents would say that all these areoutweighed by the value of dense observations of real life.Ethical questions abound. In a well-known study of gangculture the integration of the researcher collapsed when hewas asked to commit a serious crime. You may have alreadyrejected ethnographic method but do at least consider thegeneral principles as they illuminate all qualitative work.The constant rivalry between 'hard' quantitative approachesand 'soft' qualitative ones has led to the latter developing acomplex epistemology (theory of the basis of knowledge)and some highly structured ways of recording data.Ethnomethodology attempts to analyse communication, its

37


51/145


sequence and meaning. Even pauses and hesitations aretranscribed in a notation system and linked with the setting.Fascinating, but only to be tried under expert tutorialguidance.Case studyPerhaps the commonest approach but also the most abused,this is a study concerning one particular happening, orcase, examining events and facets of the focused area in ameticulous and systematic way. This latter point mustbe stressed as poor case studies often degenerate intouns tructured descriptions of randomly chosen features.Case study is often classified as 'qualitative by definition' butquantitative techniques can be used. The hypotheses usuallyemerge from the study, although a careful conceptual frame-work should be present at the start. Avoid scattering datalike confetti.A vivid realisation of structures, power, hierarchies and un-suspected variables may resu lt wh ich can be generalisable tosimilar situations depen ding on the uniqueness of the focus.However, resist the temptation to make universal theoriesout of trivia. It is probable that the researcher will have someinvolvement with the case and this should set alarm bellsringing about the danger of subjectivity.A social worker may refer to one person or one family asa case study but for a dissertation the term refers to arecognisable area such as a factory, school or departmentwith definable boundaries. There is inevitably overlap w ith

38


52/145

GETTING STARTED

other styles and case study may be termed a procedure ortechnique rather than a style.Historical researchFor most students this will mean the use of historical ordocumentary sources to illuminate a contemporary problem.It requires the same academic rigour as all other methodsand may be more difficult to carry out. If you intend to usethis method its acceptability must be checked with the tutor.Some questions to be considered include:

What is the exact problem to be investigated?Is there a hypothesis to be tested?Is it intended to discover new knowledge?What sources will be used?Can they be accessed?

For example, a student proposed a draft title 'The effects ofchanging policies on local authority planning permissionssince 1974'. Apart from the vagueness about the locus andscope of the work, it is possible but would require access toPlanning Committee minutes, Government papers, perhapseven local newspaper archives. Such a workload is certainlybeyond the scope of a single researcher.Documentary analyses may be combined with other styles ofresearch or to provide background data in case studies. Likesurveys this seems to have a great attraction for studentsw ho often, unfortunately, take it to mean the description ofthe contents of a document. There must be a closely defined

39


53/145


54/145


55/145


1. Give yourself a legitimate title, eg Mrs Alison French,MA Research Student of Anytown University.

2. Adopt a polite, professional manner in dealing with allcontacts.

3. Dress well when doing your field work: people still haveprejudices about students.

4. Word process all communications to a high standard.

5. Keep a generally formal tone; people can be suspicious ofchattiness.6. Explain to those involved what are your general intention

and purpose.

7. Offer to share results where appropriate.Your status will also be affected by the quality of yourresearch instruments. This will be returned to in a latersection.

SAMPLINGThis is potentially a very complex business with many trapsfo r the unwary. It is basically the obtaining of a manageablepart of an object or population that supposedly possesses thesame qualities as the whole. The physical world presentsfewer problems than the social one; it is easier to obtain aone gramme sample from a tonne of limestone than to get anaccurate sample of a hundred unemployed people.

42


56/145


57/145


50-year-old population; matched samples where two groupsare found as alike as possible or clusters which are groupsdefined by area or env ironm ent. Further sophisticationsmay be built in such as random sampling within a cluster.Systematic sampling is the selection of names at a giveninterva l, say, every ten th house in a street. Such a sample hasthe potential to b e very biased, for example there may be 25old people's bungalows.Even a 'given' sample as in a small case study should bedescribed and justified.Try answering the following questions about a real researchproject:

The UK Co-ordinating Committee on Cancer Researchis investigating the value of screening women for breastcancer. Their sample is 200,000 sym ptom -free wom en intheir forties, comparing those who are given mammo-grams with those w ho are not, at m any different centres,over 13 years or more.

1. Is the sample large enough?2. What kind of sample is it?3. Can you foresee any problems with the sample?

Recently, the author observed a researcher with clipboardstanding in the centre of a pedestrianised street in a smallnorthern town.She was stopping people walking down the centre of thearea and questioning them about their modes of travel andpurchases. She seemed unaware that the shoppers had seen

44


58/145


her and had pre-selected themselves into three streams; thetw o outside ones that did not want to be interviewed and thecentre one that did!It is likely that the answers from the two groups wouldbe dramatically different. Some interviewees are so eagerto please that they give the answers that they think theresearcher wants.The 'invalid sample' effect can be easily demonstrated byasking subjects for their ratings of a list of TV programmes.The list contains som e m ade-up ones and some that have notbeen broadcast for 25 years. Yes, you guessed it; thosecategories get lots of votes.The danger of a 'polluted' sample is even greater foramateur researchers as they are so committed to theirhypotheses.Another example of the risks of poor sampling can beobtained from studies such as those of drink driving. Youwill not find anyone to support the idea that d run ke n drivingis a good activity, yet it is demonstrable that in terms ofstatistical driving danger it comes a long way down the list.How does one sample drunken drivers? Those convicted?Those showing a breathaliser reading? Males? Females?What abou t the time of sam pling? 11 a.m.? 11 p.m .?

There are more, but that is enough to explain w hy amateurswould rarely tackle such studies and why they are usuallyconfined to the physical effect of alcohol on driving.

45


59/145

A / R I T I N G Y O U R D I S S E R T A T I O N

Medical research in the physical, scientific mode is oftencited as a model for us all. The author has participated, asa patient, in a so-called 'double blind' experiment to seewhich of three drugs had the greatest effect on a disease. Thedrugs were supplied by the researchers, re-packaged andlabelled to a code known only to the second group ofdoctors, a placebo included, and administered by a thirdgroup who supplied the results to ... and so it goes on!Th en, of course, there is the H aw thorne Effect (see page 37).Students should not be depressed by the problems ofsampling: improve your sample if you can; own up andexplain if you cannot.

FOLLOWING GENERAL PATTERNSYour organisation may give precise guidance about thestructure of the dissertation but is more likely to allow somefreedom. You will find it a powerful help and incentive tohave a framework of chapters or sections with indications ofthe proposed contents. This can be refined and modified asthe work proceeds.A typical dissertation will have between five and eightchapters on the general lines below .

STRUCTUR ING YOUR D ISSERTAT ION

IntroductionWhat is going to be done? Why are you doing it? Who islikely to be interested in it? W hat exactly is the hypothesis or

46


60/145


problem? What is the possible use of the research? What isthe locus and focus?Literature reviewWhat have others said, written or researched about yourtopic? What theories illuminate your topic? How does theliterature relate to yo ur research questions?Research methodologyHow will you approach the empirical work? What style andtechniques have been chosen? Why? What samples, tests,observations and measurements will be needed?Data analysisWhat data have been found? W hat is you r interpretation ofthem? Do they prove or refute an hypothesis?

ConclusionsHow can you summarise the work? Are there any actions orrecommendations to take? W as it all successful?BibliographyWhat books, journals, papers and other sources have beenreferred to throughout the work?AppendicesAre there any extra details that the specialist reader couldrefer to if necessary?The above may be modified according to the subject areaand the style of research but the general pattern holds good.

47


61/145


Some institutions prefer to split the dissertation into threesections:rationale and designmethodology and data collectionanalysis and conclusions.

Generally speaking, the smaller the number of recom-mended sections the harder students find it to organise theirwork. This is especially true where the literature review issupposedly dispersed throu gh two sections or is described bythe institution as a 'theoretical analysis'.Whatever the area - science, arts, education or social science- youwill be required to demonstrate:

extensive and relevant readingan understanding of the theories that underpin theresearchmeticulous empirical worka knowledge of academic conventionsthe ability to report effectively.

R E A D I N G F O R A D E G R E E'Reading' for a degree is the traditional expression whichmeant taking a degree and it reflects the dominance ofreading in the process. The whole dissertation should showthe evidence of wide and up-to-date use of literature from

48


62/145

GETTING S T A R T E D

the introduction to the conclusion. Some lucky people canread anywhere and in any conditions; most cannot. Findingreading time is difficult in a busy college and is even moredifficult if you have a house to clean and small children tolook after. Some simple tactics may help.Planning your timeShort, frag m en ted sessions tend to be of little use. Two hoursis more productive and an occasional complete library daywill be essential. It is necessary to plan reading just like theempirical work with careful note-making and recording. Aruthless approach will help you avoid drowning in a sea ofbooks. Never read aimlessly, hoping that something willshow up. Use the contents and indices of your possiblesources and read only directly relevant material which maybe one chapter or less.Your references may be drawn from books, journals,periodicals, newspapers or any appropriate source butremember that all subject-specific books date rapidly. Somestandard works remain almost permanently relevant, but bewary of anything more than five years old. Journals offer abetter chance of reports on more recent studies.Before travelling any long distance to a library, telephone tocheck the availability of what you are hoping to use. Booksin heavy demand may be ' f rozen' , that is on temporaryreference or on a shorter loan period than normal. Somelibraries have reciprocal arrangements with others andpermit borrowing by another college's students: if not,most can organise inter-library loans and materials f rom

49


63/145


the British Library. Periodicals may only be available in thelibrary and not for loan.Recording your readingIt is essential that for every work consulted you record inyour note book:

the nam e of the author or autho rsthe full titlethe edition, if not the firstthe publisherthe date of publicationthe ISBN numb erthe libra ry classificationpage number, if very specific.

For journa ls and periodicals it is also necessary to record thevolume and number and always the page numbers.As these are recorded it is very useful to log with the detailsthe reason why the book was consulted and a note ofsections that may be returned to later, parts that you intendto use or quote from and actual quotations that m ay featurein the literature review. If possible, handle each book onceonly. The reason for this is obvious w hen you remember thatan average m aster's thesis contains 90 references.

50


64/145


Start with the general theoretical and background books andprogress to very precise and recent work which is likely to beconcentrated in journals.Students have devised some very effective systems for colourcoding details of books using coloured inserts or dividers inloose-leaf folders; for example, red for literature review,green for research methods, blue for quotations. If you arefast and confident with a PC, equivalent systems can bedeveloped using different fonts, bold, or italics. Do notbe over complex: go for simplicity in all your systems.With luck, your first few books should set off a chainreaction as their bibliographies lead you to related work.Be selective and not over ambitious at first as the literaturem ay divert you to a different emphasis. When you feelready for a literature search there are several exciting possi-bilities.Almost every subject area has indexes to periodicals, officialpublications and reports such as the Business PeriodicalsIndex and the General Science Index. Most are available onCD-ROM and often cover the USA and Europe. These needto have a planned access as the main problem is informationoverload. Feed in the word 'Education' and you may get10,000 references! Add the additional key words ' infant ' and'deafness ' and a manageable total emerges. As with mostaspects of a dissertation the vital concept is focus.If you have access to the Internet you may find referencesand contacts, especially in American universities, but bewarned, the potential for information overload is even

51


65/145


greater than on CD-ROM. A genuine example of a searchgave the following numbers of references:Nursing 430,024+Children 30,000+Menta l 10,000+Autism 400+Staffordshire 1

We have the technologyThe death of the book has been much exaggerated. Theawesome power of the World Wide Web can revolutioniseaccess to global information but in some ways it presentsnew traps for the student. Despite its speed it can be agreat consumer of time and you must avoid unplannedsurfing.A nursing student wanted to research, on line, the effects ofbeing a premature baby on subsequent childhood illnesses.An unrefined search on the Internet, including the conceptof research, offered 5750 references under tw o categories, 24W eb sites and one news story. An examination of the firstten took 30 minutes.It quickly became apparent that many of the references hada peculiar idea of 'research' taking it to mean narrative oropinion. Many were repeated several times and the bulkwere purely commercial. Several invited us to 'Visit th e cutepreemies on the Net' - yes, really!

52


66/145


67/145

WRITING Y OUR D IS S E RT A T ION

CHECKLISTHave you:considered your overall research style?chosen and justified a definite methodology?formulated an hypothesis where appropriate?categorised your sample?thought of the total pattern of the dissertation and itsrelationship to the research questions?planned your reading programme?cleared all the above, where possible, with the tutor?

C A S E STUDIESPaul opts for a cas e studyWith some difficulty, Paul has obtained a letter ofintroduct ion as asked for by the open-cast site manager. Hehas decided that the work will be a qualitative case studywith strong possibilities of being generalisable to other,similar situations. The tutor seems unhappy that some of thequantitative techniques taught on the course will not be usedbut agrees to the outline and points Paul to the excellentselection of literature held in the department and to articlesin the journal Mining Engineer. He makes a start on thereading and plans a field visit.

54


68/145


Alison decides on a surveyAlison has decided that survey will be her principalapproach and that she will use a mixture of qualitative andquantitative styles. There m ay be a problem with the samplesizes for her first two research questions and in view of thenature of the institution the validity of the study may be indoubt. She can see the overall pattern of the dissertation butis rather taken aback that, by the time it takes her to readthe general marketing theory, a month has slipped by.

55


69/145

Te c hn i q u e sTHE TOOL BOXAll research requires techniques and instruments forthe collection of data. Even historical research needs astructured framework for the collection of details. Thewide range of possibilities includes logs, diaries, tests, tallysheets, questionnaires and interviews which further sub-divide into categories. In addition some choice exists withinthe mechanical methods of writing, film, video, tape-recording and computer disk.

As ever the real choice has constraints defined by the type ofresearch, time, cost and the nature of the data being sought.There must be careful matching of the instruments to thequestions being researched.

You may collect the data entirely yourself, from the peoplebeing studied or from a secondary source, and the actualinstruments may be devised by yourself or professionallyproduced. Remember that the latter are copyright and theiruse can involve a cost. Despite this, if it is proposed toadminister standard type tests such as personality in-ventories in psychology, you are strongly advised to considerprofessional ones as they will have been piloted, developedand modified, over a long period to a level of refinement

4

56


70/145

T E C H N I Q U E S

which is beyond the capacity of students to match in self-developed instruments.It is perfectly practicable to develop your ow n researchinstruments as long as you realise that it is more difficultthan it seems and needs a meticulous, critical process.Every time that field work of any description is undertakenit is wise to take notes of dates, names, places, the type ofwork, in addition to your results and observations. You willforget!Some instruments are highly specific to a particular taskand you should seek specialist advice on these. It mayalso be permissible to submit them in an electronic form:for example, video recordings in sports science research orstudent-devised software programs.Do not administer research instruments that require highlyskilled interpretation such as picture completions or socio-grams unless you are confident of yo ur ability.

MAKING OBSERVATIONSObservational techniques, used principally for qualitativeresearch, require a highly structured set-up - perhaps moreso as they are open to charges of subjectivity. The groupbeing studied is identified as a purposive sample by theresearcher as is the episode or behaviour to be observed.The researcher may adopt different levels of participationfrom being a complete participant to being completely

57


71/145


detached. The former can cause difficulties in using writtenlists or check boxes. It is often argued that the very presenceof an observer means that the episode being studied isdifferent from a 'normal' situation. However, this applies, indifferent degrees, to all research except physical experi-ments.Self-devised tally sheets and checklists should be clear, con-cise and unobtrusive. Figure 3 shows an example from astudent studyin g classroom behaviour.

Presumably, the student intends to define the categoriesshown. What is 'disruptive behaviour ' or 'personal activity'?What periods of time are involved? How is it possible toobserve an entire class? The sample illustrates how easy it isto invalidate an instrument and warns that what might beblindingly obvious to the researcher is not so to the reader.Making observations should always involve a two-stageprocess:

recording the 'raw' data in the note book or log sheet atthe episode sitereflecting on and considering th e collected data at a latertime.

PREPARING QUESTIONNAIRESSupervisors tend to turn pale when students say things likeTil just knock up a questionnaire' . So tempting as they canbe relatively easy to administer and reasonably cheap, but

58


72/145


73/145


they are fu l l of traps for the unwary . The f u ndamenta l r u l e isthat every respondent needs to understand what is requiredin the precise sense that the researcher meant it . Normally apilot run is used to show or confirm this fact - it rarely does.If for any reason a pilot is not possible, then the minimumacceptable is 'piloting' with a tu tor .Try to see the questionnaire from the recipient's point ofview. Are they likely to be interested in it or co-operative?How could they be encouraged? Some commercialmarketing organisations include a free draw with theirquestionnaire and stil l get low responses. Remember that alow retu rn ra te m ay have research significance and has to bereported. Y ou will be fo r tunate if you reach a 70% rate ofresponse on personally collected questionnaires and postalones may be as low as 10%. Various fo l low-up methodscan improve matters bu t cause delays and increased costand could conceivably invalidate the results by making anintended cross-sectional study into a longitudinal one, that isa study of the same populat ion repeated over a period oftime.Structuring your questionsLanguage used by the researcher must be simple, direct andappropriate to the target popu lat ion.Attempt to relate each question to your research quest ionsor hypothesis and you will find that many can be edited out.Also remember that if the scoring system for the answers iscomplex, extra hours may be required that are not budgetedfor.

60


74/145

T E C H N I Q U E S

People hate confusing questions that demand different stylesof response from one part to the next: ticks followed bycircles followed by wr iting . Also ir ritating are changes in theformat of questions. It is better to have a 'house style' for thewhole questionnaire.Closed-end questionnaires are easier to administer and toscore but can produce bland or sterile answers for example:

Which of the following Leisure Centre facilities haveyou actually used? (tick)

swimming poolmultigymsports halloutdoor fieldnone of these

Simple m ultiple choice ques tions almost always need a 'noneof these' or 'don't know' or 'other' category.Other scales may be used but some target populations findthese difficult to cope with; for example, Likert scales:'The Leisure Centre provides good facilities for motherswith small children'Tick the statement that best describes your opinion:1 strongly disagree 2 disagree 3 neither agree nordisagree 4 agree 5 strong ly agree

This type of question does not cover respondents who arenot sure or who have no opinion.

61


75/145

WR I T I N G Y O U R D I S S E R T A T I O N

Scales that require complex instructions only work well withtargets who are involved and committed to the research, forexample semantic differential:

Place a vertical mark on the line between the two wordsthat m ay describe your attitude towards some featuresof the Leisure Centre. For example, if you feel that theatmosphere is friendly put your mark near to that word,friendly ...I . . . I...I . . . I...I . . . I...I . . . I unfriendly

In this case it seems that the information obtained does notmerit the complexity of the instrument and again there is noscope for the respondent who thinks that the atmosphere isdownright hostile!Open-ended questions offer the respondent a variety ofresponses:

'What is your opinion of the facilities for mothers withsmall children?'

Such a question m ay produce deep, insightful answersbut equally m ay result in a difficult to read 'All right' or'OK'. Single researchers could easily introduce a bias at theanalysis stage.The unsatisfactory wording of the semantic differentialquestion above illustrates how slippery words in questionscan be. Vagueness breeds vagueness:

'Are you acquainted with the facilities?''Do you generally use the swimming pool?'

62


76/145

TECHNIQUES

What does 'acquainted' mean? Have heard about them?Used them? 'Generally' is a very common word in students'questionnaires which means very little. It is very easy tobuild in disastrous ambiguities:

'Do you enjoy travelling via the Channel Tunnel andthe ferr ies? 'Answer: Well, yes and no sometimes.

Questions should not lead the respondent by assuming aparticular stance:

'Are the severe government cuts responsible for . . . ?'Questionnaires, more than any other instrument, need theapplication of the 'so what?' test. You need constantly torelate every section to your research questions. Question-naires seen recently seem to be full of questions like:

'Do you think that tutors should take an interest instudents?''Do you agree that seminar rooms should be warm andcomfortable? '

These tw o were followed by a Likert scale ranging from'strongly agree' to 'strongly disagree'. If the researchquestion was, 'What factors influence drop-out rates fromuniversity courses?' the data collected will have littlerelevance to providing an answer that was not blindinglyobvious already.

63


77/145


Despite all these strictures a good questionnaire can beinvaluable for producing large amounts of valid, handleabledata with a high degree of objectivity.Introducing you r qu estionnaireNever distribute unheaded, unexplained questionnaires. Asa m inimum the heading should state the origin and purposeof the work; more extensive questionnaires should have aletter attached in which the overall purpose of the research isoutlined, brief details of availability of results are given andofficial contacts are named.

CONDUCTING INTERVIEWSInterviewing is a method of collecting data that can standon its own or be a follow-up process to another method.Interviews should never be random as they demand a heavyinvestment of time. One comparatively short interview cantake an hour to organise, an hour to carry out and half anhour to reflect on.Interviews can also save a lot of time if used as preparationfor the structure of other research. The researcher's per-ception of reality may not correspond with that of the targetsubjects: a pilot interview c

Writing Your Dissertation How to Plan, Prepare and Present Successful Work

Documents