Research Questions and Objectives The purpose 'pi research ii lu tlin-iier imuvn :« qucitttMi through the Introduction The Uic of research questions it a neglected aspect in the design and conduct of social research. Tnh it surpminp, giver. :hii the fundamental purpose of social research is to provide new knowledge about the social world, to answtft piadotes about what the social worid is tike and how it works, and io rind WWJTS to solve problems and bring about change, la my vtcw, formulating research qutt&MMS is the most critical and, perhaps, the most tMicuh part of any research design. Ir is only through tog of ■S»*h questions thai choices about tftt focus and direction n: research can be mads, that its boundaries can be clearly delimited, that manageability can be achieved and that a successful outcome can be anticipated. Establishing research questions alio makes ft possible to select research >iratrgiei and methods with confidence. In other words, a resfVch project & fcriift on the foundation of its research qUtStio)i$. However, fccntns; these- questions dear and precise requires considerable diou'ht iud sometimes some preliminary investigation. This chapter discusses: • three main types of research questions; • the functions of research eucstiens; • how to develop and rcl'me research questions; • the rdarionship between research questions and .-.ypwheics, *n± the tWctions of lbs latLct: • how research questions can provide a jgitide and framework for the review of the 'iteraiLte; ' • the nature and range o: research objectives ilwt can he pursue^ and ■ the relationship between research objectives and research questions. Research QtrestibrK 59 The aim of the chapter is not only to argue that research question; are necessary, bu: also that. ;too:; research needs high-quality quest ions. A rare attempt to deal widi the Issue of the quality of research questions has beets undertaken by Campbell et al. [jJfJS). They reviewed articles in five journals in psychology. organizational behaviour and management, taking a two-year period for each voumai. A lis: of the research questions was compile*] and chen iliey surveyed researchers in the fields covered to see what questions they thought should be asked. Their am. w.ls to find ^gps in research and to establish priorities for future research. The important thing about their work is that they focused on research questions. Research Questions Gon\eiukmal wisdom suggests that research should be guided by one or more hypotheses. In order to e;ct started on a research project, 111i^ view advocates that the researcher should, first, sdee: a research j.iroblcni, second, state one or more hypotheses to be tested, and, third, measure and correlate the variables related to the concepts in the hypotheses. However, this procedure is only relevant to quantitative research conducted witlun the tietiurJhv research strategy. W'iu.i there is a role tor hypotheses in particular kind* of research, they neither provide' the foundation for a research design nor are they very useful for defining the focus ind direction nl a research project. In fact, the ritual of fwranlaiuit; mid tts'fiuj hypotheses can lead to uimcccssft*} and unhelpful timidities in the way in wbic.1 research is conducted. In some kinds ol research, it is impossible or uraoeoessAty to set (put with hypCithcscs.A imich more useful procedure is 10 establish otic or nt,ore research questions. lew textbooks on research method1; jdvc much attemfon to the formulation of research questions, and some ignore this vital pan: of the research process entirely. Exceptions can be found in sonic recent texts on research methods, for example Yin {\99% llednck c-t al. (t993), Miics and Hubeiii-.an (1994), CYeawe.l (I ?$M, 1993), Marshall and Kossm.ni 11 9$$}, Dlaxier et al. (1996), Mason ;i<>5t,j, Not-man (1997) and hlick £19*3). It is interesting :o note that these books- arc concerned with i,uahta:ive research methods or include a sipniixaiit discussion of them. Metiman, for example, has regarded research qjestions as the brtdee between the research topic and hypotheses; he has a'.so Giscusscd tcchr.iq'jcs for narrowing a toptc into a research question. Y:n has liaked research questions to types of research designs (or what he lias called strategies}: exj-eriitien:, survey, archival analysts, history and case study. Flick has argued for 1 he importance of research qaeitions in-quaJrative research. ■■\ .^r.rr.i! step. f:oiit«J wirh this zir^bfem net oaly at the begLnrur.s;, whtt rhc stuav or ;;>e project is COCeejfflisliled, but in scvera; pfi35C; of *' prceess; in o)r»ccptua:rdr.g rhe tewarcn dcsien, ir. enrctui* the Lisle, in sc-ecring die Cas=s acid in colicctiili dt:ia. [Flick lWfi: 4?) Fiestai^i y<*=*>»OT8 s«>a wf^^UiEi Lr:sw=l| has offered adrice 03 bow to formulate re^arch questions in quannca-tivc and qualitative research '15941 and In five rradit-.oss cf qualitative research [1998). Mason set her discussion o: research questions tn ike cor.rext of intellectual iiu/iles thai seek some kind of explanation. Thcsj puzzles take a variety of forms, depending on the on:olo);ic2l arid rp-.stenui'ogicai positions adopted by the theoretical and inttUccmal traditions i'roin within which they emerge. [Titcllcctjil piwzlei. then, wiJ can-ji:r. different sett irf c-ntoJovpci! ar.d cftstcmclo-g;ca5 assamptions and prescription;, and m) surest ' BJttfl to move yuu Irnm your broad te.wreh inter^n to yojr tpecjjte research io>ui and project, and ihirciou unit iir,pvit;.»ittc ca4a;t be ovc:i«:cd. Rcscjili (.;iiritio;ii, then, arc ihi>«- qjeHiems :o w:»*h you a* rtitttrebe really wart :o i,;n--w tt:r answr:*, aSC m that jca*e they ;rc :ac iWeud exrriijiaii of your intcllt a Whac types of perij^le arc •nvol-.'cd? o 'Ot'hii tharacieristic kr.cwted-^e, b;licfs, valurs and aciiracies do rh£| hold? Contexts Core elements Tapkrfproblem ) JL Choi Research questions and objtirtivet How i;u«rtiDtit Panhort. good deAC.'ip-tion is a vital part of social rrscarca. Some writers have proposed more ihan three types of research questions. Yin ', l''^_>), for example, has discussed seven types; 'who'.'what', "where",'haw many', 'how much', "how" ar.d 'why'. However, he doe*, acknowledge that 'who*, "where', 'how many' and 'how much' questions arc different forms of a "what* question. Hl.ixtcr et a', have suggested five types of questions; 'hyw'. 'who', 'what*, 'when' and 'why'. Similarly, the iirst fottt of their question* can all be transposed into 'wb.it' questions: 'what individuals' in "what places', at 'what time', m 'what numbers or quantities* and at 'what ways'. j\ different .ippr&ach 1o research questions can ttt tout.d in Hedrick Ct '1993 23-31). They have identified lour types of research questions that ate relevant :r> applied research: descriptive, normative, correlative and impact. Mat shall and Rossmati (1995) have classified lescareh question^ as theoretical, as focusing rtn particular populations and as btin^ site-specific. These Categories relate to the context in whii-h they ate examined. Qucsitous. may br theoirlical ones, v. etch can be restarchtc in any one of a ucmbc: of different dtts or with diJlricta satariri. Or lijty nuy be (oeusvrd on a pj:::cuUr fO|jjl,iiiLiii o: l'.isj uf individuals; these it-o .iar. be .iLUiid Li -..ui^ttt plates. Kricily, the qiwsiicins any be '.itc-ipecii.i: "sccauie of the aniqjr«ejs of a .specific program or "rganir.aituri. (Marshal aitJ "los-miri IMS: I?i Throughout the book, I shall dUcuss only the three type> or research q\iesrions. 'what', 'why' and 'how'. The process of de\elopin{= rrscr-rct qaesOOQS wili Llesit-ably produce a range of question wording simdErio that discussed by Yin (IS93J. However, 1 believe rhe discipline of reducing all questions to these three types heljis ro make the links bcrwecn research questions and research objectives clear The Purpose of Research Questions Research questions are nccce;: ro eeiio.c the nature and scope of the research. &y selecting questions, and paying attention to their wording, i: is possible to deosr-iTune what is to be studied, andi to some extent. ::yv.' it will be studied. Tne way a particiilai research question is worded can have a significant influence ort liow ruoeh and what kind of research activity will b« required. Let us return to the four research topics discussed in chapter 2 and examine some possible research questions for each one. Lfivtrotimetaa't Wortdmcws and Behaviour ttfnoiig Studcnis and Residents 1 To what extent do students and residents bold different environmental worlil-v.ews.' 2 To what extent is environmentally responsible behaviour practised? 3 what is the level and type pi involvement in environmental movements? 4 To what extent, and in what ways, is environmental behaviour related to environment a I worl d v iews ? 5 In what ways and to what extent wilt environmental worldvirws and behaviour ehaiy,c over the next five years? Aa these arc all 'what' questinns, the st.idy will have only desMpUv* r;b|ectivrs. 1? seeks to describe the distributions of environmental wurldviews and behaviour in these populations, and the paiierr. of the relationship berwetr. these variables, now airul in the future. Af>t? aid Ewirtmvitittntisni: A 7«:j,' of' Co*»ftefing Hypaibcsei I To what extern is ag* related ro environmental worldvicw* and environmental behaviour? 1 U there ate :.i:-i:Hi-.'.ir.t rrlaihnisliips, what arc the- Jo-ir..-. nenUii Aciri'isis 1 In what range and types of behaviour do en viror. men tally responsible individuals cr.j;age? 1 ^C'hy do these people Set responsibly towards rhe environment? 3 "Why do some o; these people manage ic sustain this behaviour? 4 f inw can the incidence ct this type uf Ivehav-our be increased? 44 fl;;p3fcfi Qterticrrs end OfcyeOrres Now come to a combination GU all three types 61 research questions. The study-seeks to describe znvironrnerjtslly re spun able behaviour, ani tien to explain why people engage in 2nd manage to sustain that behaviour. Then uorncs the sting in the ;ail - how to get more people to engage in this behaviour. It will be unlikely that this study can do anything mere than coin: in the direction of possible answers to this last question, using the answers to questions 2 ar.d 3. But it could also suggest ideas for further research to pursue it- iSe: the Appendix for examples o: different and more compicx sets of research questions.) It is important to recognize that while it is highly desirabie to prodtxc a wcli-formulatcd set oi research uuesiions as part of an integrated research proposal or design, this may not always ks possible without sonic preliminary research heiu^ undertaken. Ir. addition, what is discovered m the process wF undertaking the research :s likely to require a review of thr research questions from tine to time. >5o research design can completely anticipate how a rcser.rrh project will r%olvc. 1'. may turn out that some research questions cannot be answered because i: :s not possible lo obtain :hc necessary data. What the researcher assented, or was k\l to believe, about the availability of or acccssto the necessary data may turn out to be wror.fc, Consequently, the design may frcmre some revision, and part nt this may involve a change to one or more research mictions, t lence, while it is necessary to be as clear a> possible about the scope and direction o: trie research at the l)(.-i;iiitiiiiy;, Whaf tht researcher learns in the course cf undertaking the research tn.iy necessitate some changes. This it simply the nature of research in any discipline. Ki-scatdi projects differ Hi the extent to which it is possible to he able :o produce precise research ljjcsuuus. This, is certainly true of exploratory research, the aim of which can be to provide information to assist tn the development of re scar en questions. It might also lie .iri'.ued thai some studies that us; qualitative ui Ltli[iuL'.raphic research methods involve (he researcher :n .1 learning process. In these- cases, the research questions may evolve in die ceut>e of the c:>e.irt.h. However, even this kind of research requires catefui consiJccacion of scope and direction a; the leaning in order to ensure that it will be manageable and will have a high probabilky o; successful completion. The developmental nature of a research design should no-, be u*d as -in exc .ise fur avoiding the effort required to formulate appropriate research questions. A con-mon feature cf the research process is for the researcher to be deflected or c.ist:;*:tad from their original intentions. Many nittuences may be a: work: • encountering new ideas, for example in published research, in conference papers' or urcseriiariuns, in previously urlmnLiar theory, o: in th. mcdia; • discusNion with collfcs.g-25: • chanfiinj.; academic fashions; « changing political agendas; and, nor; particularly, • learning that ta.e rend regularly to keep the focus cf the research dc^ Developing and Refining Research Questions The process of developing a sec of research questions can be the most claalkmJL part of any research project. This is particularly the case when the research* initiates the project, as is the case in much academic and postgraduair re>carc.i b the social sciences. However, die problem still exists in research that is commit: sioncJ by someone else fur problem-solving or p61iey-rela*ei] piirpriscs. Organn| tiens or group-, that commission research are very oltcii vacuc alxmt v.Har ituy-i want done, .ir.d usually need some assistance to scarify :hc research v.nesiic-ns ^ obiectivrj.s. It i5 very rare for:i researcher t« comtneriCc a project with clearly kumclned research questions tilrcady provided, This might occur where a researcher hail' lointd a research programme in wlnth the research quesiiuns have ajready been established, or if a researcher is taking up question* posed in previous research. Ilc.vrvcr. i: is in.ich :nnrc comutoii in ine stJCial st-icjicns for rejearchrrs ta approach n topic we field in which previous reviMn h it limited, rn i:i which previous rc^eircli has uted :m approach different from the one the rcsearchet wiriu to tise, nr considers tu be appropriate:. All researchers have to devise their own way of dcvebpniL'. research qucsticns. What I offer here is a process thac I have iibtd myself and found uj work successfully with many postgraduate ntudcms. llwertrch question can he simiu!-atcd in many ways: iron: casual observation of possible tegul.ir:l;es; from prrvioul research from dieory; from, reports in the nieilia; nr from discussion* wttb collMgycs, The source is not realty nuportan:. What it is usually necessary ii> id bring svnie order into a range of loosely connected ideas about what should 1« researched. N'cuman (1997: 122) has offered n stmdar set oi wxhniquw. J Wrttn tiuuftt ..wrv .jwes.'fuc yo« can tbinii of Lei both the results of your readingrtnd irua^inarion run riot torn reasonable period ot rime, and record even question that occurs ro yon. Brain-storming sessions, rm your own or with others, may stimulate the process. Note down cuesrions when they occur to you, wherever and for whatever reason. The list Will include all kinds of L]uesTiciib; som* will be seeking descriptions, some explanations, sonic wi.l be concerned with ■u-jor., anc so oa. There :s no need to try ;o cchieve any order or consistency ir the lis:; simply record die filiations as they arise, O-ie question will usually itinmlate other questions; they shou.d ail be reccrdrc. This activity may produce a very lout; list, sometimes many scores of rjjesnons. Th; purpose in doing i'^s is to try to expose nil the ideas that you havp on th; topic, particularly thosr that may be taken for granted and winch later you wish you had"been rally'swarc of be the d^ign stage. No quesiicn shoJiJ be censored, even if it may SSsin to be murginal, outrageous, or iinprachcal. ňesccrrh Qjesiiors 2 Review the list ofquesuunf Once ywu arc satisfied that yon have pretty well exhausted all the ideas you have on the :cpic, jmou should review your list. There arc a number o: strategies for dDini-: tHjs. • Group the questions under similar themes or topics, if such exist in your list. This is likely to reveal overlaps between questions which will make i: possible to eliminate some ar.d to consolidate others. Pan oi this consolidation can be achieved by developing a single, general or abstract question tna: summarizes a gM>utp oi ruore specific questions. » • Set nide questions ihat seem to he outside your raaiu area(s) of interest, tint ;iri- ::)o outrageous, or that sren to take you in dirctiions that may be to:? aitficult SÍ WO demanding: to deal w;ih. Vol can always review these questions bter if you decide to change the dircctir.in of the research. 3 Seftatttt* 'whn:'. 'wW jijj 'how' fyteítítitíi WWltn each group of question-, bej;i'» ip identify those that appear to be 'what', 'why' and 'how* questions. Of course, tone vtudiei raav he concerned ultimately with only -jtie type of question, for example one or more "whať questions, or just a 'why' qarsi.on. The wording of "what', 'way' ar.il "how" qurMions requires irry careful consideration, as tlie way a question is staled initial.)' car. be tkxrpt-.vc: Vbať ar.d •why' ttiiriitior.s cín U%tn with 'How', and 'how' question* -an lie&m wr.n 'VX'rur'. l:or cxanqw: 'How arc environmental kgihniiOuf and environmental world'.-few related;' This needs be transposed into a descriptive question, as: 'Whit i* the relm innslup between environmental brhaviuur and environmental worldvicw?' or 'To what extent, and in what ways, is environmental behaviour related to environmental woildvicw?' Die question, 'I low do some people manage to behave tn an environmentally responsible way?' needs to be transposed into an explanatory question: "Why do these people act responsibly towards the environment?' The question, 'What can be done to increase the incidence o: environmentally responsible behaviour?' needs to be transposed into an intervention question: 'How can the incidence of environmentally responsible behaviour be increased?' Make sure each question is worded as clearly and as simply as patele and that each one can be identified imaabijjuously as a *whať, 'why* or'how' question. Complex questions may need to be broken down into a series of questions. For example:, the question, 'Whar is the incidence oi student plagiarism:' would be better broken down into at least two questions: 'What has been the extent of detected student piagia:isni over the ras: five years?' and 'In what types zi plagiarism have students engaged?' (see the Appendix". 4 Expom JHutttpttL'tts Cheek each r.u-~sri:an to see what it žisumes. Many questions, particularly "why' cajettiruic. presuppose ether questions. It is irr.por:-ant id expose the 'what' question that must ce answered b;fore a 'why' question can be asked, or, perhaps, even formulated. 'How'' questions may prcrjptJttt? ro:n 'witat' saz. particularly, 'why' questions. A rcsearcn project iv.ay need to examine ill three types of questions. Rather than reducing the number pi goestipas as the hit, this part of tiie process may add further qucsrions- ■. .i-,'.«il.ible. "laetc is in inevitable icudeticy to try to do too mncli; tin* questions for rhc topic on studen: plagiarism ure n good example [see die Appendix). Therefore, tt ,s nj-.'ii,nble 31 ti'iis stape to redticr this project to what may appear ro be n:i cxtTctnciy limitec o: even tribal set ot questions. Such innocent-lookiuf* queiiinns usually liavt otJier qii:siio:is lurkinj! i:i chctr shadows. A uier.iplkir to illoscair tiie need to narrow die Joi us oi the pri>)cct :.i tins star.c car k- taken Irof.i the shape wl .u. :u»-.ii|.;::iss, the :.ticieiil met land ol mtasurtnt: time fuje figure 3.2). At die lacginninj; ot the des:j;n of a' project, our jdras arr usually wide-rariynj> and scattered, as represented by the broad top of the hourglass. What we Heed to do is to narrow, cmisoLdite ami (iku:-. dn-x. .d.a. So tli;,r lUey can easily pass through rhe usck of the hourglass. O-.Kt we Itave achieved this, and the research cotimiíiicex, w: aic: likely ic iuid thai the idc^is and qoeittíons Dcyiin io expand, and by the ei^d of the prnjeci they ms,y hawgYuwii tr* the si/e of the bast; of the hourglass. Wh.it :uust ensure is chat the si^e y qm$tw>ffi Once the list of questions has bcren reduced lo what appears to be a nanaj!rabk sti, further work car. be done Oil them. U may be ttScfui to separate the qucsdons into uvo broad catígoiies. major questions subsidiary questions.1 Matot research questions :ane these that wJI form the core oi tbc research project, the key l^iísiío:is rhat ňre to be answered. They may also be stated more abstractly íhun sonic of the oilier questirnví. Research projects rraay have only ctie major research queition, per baps a Vhať 1 Hejriek etal. (1993) hívescfificsiíii a siinil.nrdi-vision benk'eca 'pranary' »ik1 'iniiftfiiriiiEt" taatú. ques'.ioai. ResecrrJr Qjesíions Okjĺcťneš question. However, most are likely to have a combinanon c-t major questions: 'whar Questions and a 'why' question, or a ňéi of "what", \vhy' and 'how' questions. About five or six major research qucstior.s ;s procablv ni^-rc tain enough for any project. Subsidiary questions will include those th.u deal v.ith background information or issues tha: pre presupposed by one or more major questions thai, while bslnt necessary, arc net absolutely central to the project. Here is an example or a set of major and subsidiary questions. Major research qmstion • To what extent is environmentally rcspor.sirle behaviour practised? Si-íhsiiHízr.' research questiuns • What proportion oí residents regular;)' recycle household waste products? • What proportion of residents avoid buying cnsironmcatally damaging products? • What proportion of uraversr.y students arc actively involved in environr.tent.il groups? In this example, the subsidiary questions can be used to specif) c nebúria or environmental behaviour inc thus focus the snidy. 7 Is tacU (tucsimo ncctoartt As your set of questions begins to take shape, yon need to subject t hero to critical scrutiny by asking of each question: *Why am I taking this question?" 'Is it necessary?' "Why do I want tu know this?"' what will I Jo with the rcsiilts from it?' "Kow docs it relate :o ether questions?* 'Is it researchablcr" 'Can 1 manage all thr>c questions?* This process needs to be taken very seriously and no: glossed over quickly. I: is very easy to include questions because "that wouid be ;titercsti.ig :o explore', '1 would really like to know about that". This critical examination needs :o be ruthless. A common mistake in drafting research, caiestii>n> is W contuse them with questions used to elicit iniorn.atio:: Írom respondent* or pamcipanrs, for example interview questions, or questions that would go m:o J questionnaire. Research questions are what you wan: the research project to answer. Questions you ask respondents tan provide the basis for answering rexcarch questions, but tnett style and stops arc very different. A wide variety o: data iriay contribute to die HiywCfinjJ of eh y research question. Vany postgraduate studení* seen: to have a desire to do the definitive piece o: research hi tlxirtopic. Even if the rese.mh is for a Ph.D., and even it the degree is based entirely or almost entirely on a thesis, the research aiia; be limited at d focused in order to be manageable. Thera ;S ;u»t so much that one person can co within the time limits prescribed, and this can only make a small COBWibutwn to knowledge. While some students may be quite pragmatic about doing just what is necessary to qualify for a higher degree, many appear tc have a strong need to be seen to be making a major comriburion to knowledge. This ii nqf only an unrcali*: ic expectation for a 1 ally research- based Ph-Du; it is impossible in research for ar.y other kind of postgraduate degree. '1 he problem is most acute for students 1 tiff', p 1 !.í(íŕ- m ■■•".V(t w undertaking a coursewurk (taught] masttr'; degree in .vhicb. there is a m' trLesis.''diiscrta:ian/pro;cct component. Because of its limited duration, surf)9*! research project is very difficult to design. In short, the number raid nature of the questions selected l^as got to reflect id available resources. This is the stags at which, the scope of the project is dj ' nunecV and bad decisions can produce serious pre hi cms iatcr. Influences on the Choice of Research Questions The choice of research questions and objectives can be subject »q the samj influences as those affecting the choice &t tftt topic itself. ReseatLh quesac3 may be developed to satisfy personal or academic curiosity as wrh a% tu andress iwat problems. They may be inspired by the work of a particular social tlieoria ot by the results uf previous research. In addition, they may be influenced k t variety of audiences. Tlic latter is particularly the case in applied i escmr, wriite someone other than the researcher sets the agenda. In the end, however. ||K of formulating research questions, and ensuring '.bt they form a cor.sistenU set, lie* With the researcher (sec 'Inflmuicrs on the Choice of Topic' in chapter 2 above). Research Questions and Hypotheses 1; is a on:mon view thai jirjcital resentih should be directed by one or matt hypotheses. Howcvtr, in .^onie types of lescareh it is impossible or inapprop iate CQ tot on: with hypotheses. In the types of research in which hypo:hews are considered co be essential, it is not always clear what rhe role of these liypotheseJ is ot where I hey are to come from. In some tradiuote, of teseftrch, il h expened that hypotheses will he stated very precisely, in the null and directional forms, to. tacihtate stacistical testing. In oilier traditions, hypotheses are stated much mare loosely, and their acceptance or rejection is a matter of general evidence ..nc rhetoric rather than tests of significance. In practice, hypotheses arc drawn froti a variety of sources, such as hunches or intuinon, previous research, dii:ursi« argument and carefully formulated theories. While the latter is advocated in seme traditions (see the discussion of the deductive research strategy in chapter 41. thefc source is frequently va^uc and their purpose unclear. Lundherg's early [1942) textbook On social research provides a classic?! vie*' °r the role of hypotheses, lie argued that there are four steps b "rhe sciecta-" method': the fotmuiaton of a working hypothesi.s, the observation and recording of data, the-classification and organisation of the data collected, «nd the produc-cioji of general iia: ions that apply under given conditions. Li lhis context, lanidbeifi d;fmed a hypothesis as Ha tentative generalisation, the validity of which remains to be tested. In its most elementary stages, the hypothesis may he any hunch, goc^ imaginf.tive idea or intnitioit whatsoever which becomes the basis for action 01 bvcsiigaME' (t942i % Tliisviewofa hypothesis simply requites the researche* to have a gyes* at what they dilak the dain might reveal, an J -.hen proceed to see fieíea.xr. Qu es tons and Qůy.-ct/ti íteiúrdt Questions ií ii is the case. So convenooiial has this view become that the novice res: archer fuels compelled to rttaice such guesses, even ii it makes no sense to do io; one feels naked without 5 hypothesis for f. fig leaf. Tnc fear o! not being able to 'prove' their hypothesis hangs tike tnc sword of Damocles over the novice's head; guessing the wrong hypothesis, or the -wrong versioa of it, can be regarded as a disaster. The stress id this tradition of research is cn having a hypothesis not alwavs 0:1 where it tomes from, what it might be connected :o, and what purpose íl serves. It ;s not uncommon to invent such hypotheses after the research has been completed.* Some writers conIIate hypotheses and research questions: 'Wc do research to get answers to questions. Therefore, to do research, wc must stan with a research question rJaat can be answered. Tíiií qtrt^bn is usually stated as a hypothesis - an idc.-,, ii prediction, citable of beinj; disptoven' [Mitchell and Jollcy í9S2i t.5!. I rtc confusion i^ further con-.poui-.dcd in iheir ^itw thru hypotheses can be deduced trom theories, and (fiat theories tan be expressed as a series or hypotbcs.es. From this, it is iiííficult to know what a revrarch question is and what rote it is supposed to play. It ismy view (hat kypull.-a-ci arc t/llftatittt pjytVtti Mi '»8f'OHO, iomt'.imei. ■Ikw' rvscsn-li .iitcsnrtm. Iney are our best gitcssesat the answers, nut they arc not appropriate lor 'what" questions. There w ttcdc point in hazarding gjcws at a possible ft«e :.f aiUitv Rrscareh will produce an answer to a 'what* question in due course, ntd no amount of guessing about what will be tourtói is of any assistance; it might cvcti prri adice ilie answer. Thetc.'crc- hypotheses shuuld be reserved for the role of tentative answers to 'why' and 'how* questions, ar.d paiticjbrly "why* questions. While it may not always be possible to produce a hypothesis tor such research qarstions, to do so is to k»vc research a much clearer sense of direction; decisions about wiiat data to gather, and how to analyse them, are caiicr in make. 1 lowcver, it is important to note tkac some t:ad;tiuns o: research that are conferred with 'why* questions may not set out Willi hypotheses. In grounded theory, lor cx.irrpJe, hypotheses are proposed in resppiuc to the patterns in :i:e accumulating data, and they will be tested in a continuing tri.il and error process, bein>j refined and, perhaps, discarded along the way. A central issue tiiat researchers ennfrcm at the stage of formulating research questions and hypotheses (if required! :s whar concept* to use and how ro define tbem. How this as handled will dep=r.d largely on the particular research strategy or strategies, and theories or theoretical perspectives, adopted. Tttis issue will be introduced in the next jUSlion and will be discussed Ji more detail in the early par: of chapter 5. * I was i vie ran ol ibis kind oithii 11 š wliis ucc:nskir.g tms-tJ ny tu«:i aVgrw txí. ir.the 1962;.Thi iesii rqicitfri ever lairnf irj>ci!>cse=. znaíi of wttttfc vtíc ccíy ver» !oce-^li •a'ai rtponcá in ííK^i * ^a>" thai -I appcirrc ihsi S t^řn /ui^ci ťy ^hc^íí nypoď.cscs anJ "-"J ijitctullíJ/ :tj:e>j t'ir.i.. Tlir vfcaie pirforrunvc nota.cr more-ian nno. ihMl was;c :r s st.^t.i i: ±t iiKjp!:r»t ilJin'M. Vha: I ixiS.y errJi'J ují £ cle;r x; .ii rLiiiich qT.-eiriorí, tjJ *»í x Ac icxzbocin jJí-soíhíc d.r need I i iuch čtirsiiotií; oOC hid to have h yftirioes :cr tif scMsftcb :c be azcíptijáe. la ale iiari> or research report. The firsi may be pan t)f the introductory chapter; I he .'.is-t r.vo in.iy appear in a ineti'itKlolo^y mid metlmds cliapicr; and tiie fourth may be pjrt ot a discussion on (he choice of research strategy or strategics. It is the second and third, on previous cese'arch and thsory, that :ire pam.ulaiiy r the research questions. A m;s|or dileiuuui in any research project is to establish what litcnr.uic to ccview - what literature is relevant, This can be a daunting and confusing usk. particularly for novice rcseatcheis. I luvn uhservrd many students siteitd.ng ar .•KctsMVf ..in mii t of titn-L- reading r.u'iet aimlessly. Some will not really he satisfied until they have read 'cvcrythm.i'', but ;lir problem is to know what to iuclu.it- n 'e^er, ■thing". One solution ro this problem is ui use the reieatch questions to guide mill Structure the review of orevious research and re leva nr. rheory Eiiwh uuestion i:ju Li used to pur a boundary around a body of literal mi, he it theory, published r;*earch or reports. Th?-i)>n ui the utgr-crtw.C rct.-mjj if to aitttc&U what- ib&Stetc. of knowledge is w'dh reip.tci ti> each research question, or t&QHp ftf QW&ttQttg, If hypotheses arc used, they should have some connection with this literature. In some cases it may be possible to dtrive such an ar.swei from existing theory or i: may be necessary to consiruct a new theory for the purpose, hi wc shalJ see, witfan the deducivj? research strategy, the development of a theory from which a hypotheses or hypotheses cso he deduced is an essential ?ar: of answering 'why' questions. In the tctroducsive research strategy, rhe literature review may provide some assistance in the construction cf hypothetical explanatory models. When the abdnctive research strategy is used for theory generation, hypotheses are fin integral part of the coaiinuing process of data collection and analysis, of observation, reflection, hypor-;ies;?.it:g and tes'.hig. However, aavocates of this strategy usually argue that research should not begin with hypoihesM, 72 Research QueiSons and OSlieHrvfcs Research Objectives In contrast to lite researcher's personal motives aad goals for undertaking a particular research project, research objectives are concerned with the types o: knowledge to be produced. Social research can have a number of objectives ranging from relatively simple to very complex, and encompassing both basic and applied resca rch. Reseat ch cbj c etivestnclude expio.'Mion. description, expianaHbk, HRder-stjnding, prediction, change, evaluation, and impact csfffiwi-' I see figure 3.1 J, Arc-search project can pursue just oneof these objectives or, perhaps, a numberof them ih sequence. l:or example, a Study may be purely descriptive, or it might begin with a descriptive stage and then ptocced to exportation and then to charge. Basic research focuses on the fits- rive objectives, exploration, description, explanation, understanding, and prediction, bui particularly description, explanation and understanding. While applied research may include some of these "basic" objectives, i: is particu.arly concerned with change, evaluation itra impact assessment. Typos of Objectives Brjsic Research To cxphre is to attempt to develop an initial, ruugli description or, possibly, an understanding of some social phenomenon. To describe is to provide a detailed account or the precise measurement and reporting of the charaeunstics of some population, group or phenomenon, including establishing rcgulariiir.-. To explain is to establish the elements, factors or mechanisms that ate responsible for producing the slate of or regularities in a social phenomenon. To irtdenfatid is ta establish reasons for particular social action, the occurrence of an event or the course of a social episode, these reasons hems derived Ironi the ones given by social actors. To predict is to use some estabtithed understanding or exploitation of a phenomenon to postulate certain outcomes under pan:cu_ar conditions. Applied Research To change is to intervene in a social situation by manipulating some aspects ut it, or to assist the parrjeipants in doinu; so, preferably cn the basis of established understanding or explanation. To evaluate is to monitor social intetventioa programme* to assess whether they have achieved their desired outcome*, and to assise with problem-solving ar.d policy-making- To assesssocial impdeti is to identify the likely social and cultural consequences p,f planned projects, technological change or policy actions on social structures social processes and/or people. In case you might be wonderiru; why comparison, is not included as a trsearch objective, 1 regard it either as a form of description or as a technique for arriving at explanation or understanding, i.e. for theory generation or testing. In facq_ comparison is one of ihc best methods for generating thenry, :iS ;s evident in grounded theory (S'trfajuss and CTorbin 15W9). As such, it is noi an objective but it can be a means for achieving such objectives. Therefore, a list of objectives should not include statements like "To compare the environment,"!, attitudes of ui ivcrssw students and logging contractors'. A research project might set out to describe the attitudes of each group, and to try to explain why they hold particular attitudes, but a comparison of then- altitudes can he pan of either of these chicctives. F-xplotiUton lixploratttry research is necessary when very little is known about the topic being investigated, or about the context in which the research is to be conducted, Perhaps the topic has ncvci |>ceii investigated before, or nem in that particular context. ISasic demographic charm te.ristxs rif .i gfOUf ci people, ON Him - aspects ol their hekaviour Of social relationships, may need to be linnvn h order to design the study. The relevance of particular research questions, or the feasibility of usin^; certain merhods of tlatn gathering* may also need to be explored. Essentially, exploratory research is used lo gcr a better idea of wnat is going an and how it might be researched. While exploratory research is usually conducted at the beginn n.< 01 i,i..-.nch project, it may nlso be necessary at other stages to provide inlor;ii.itio:i lei . usual design decisions, to overcome sir. unexpected problem, to better understand an unanticipated finding, or to establish which avenues of explanation would be worthwhile pursuing. The methods used to conduct exploratory research need to be flexible and do not need to be as rignrous as tho.se used to pursue other objectives. The researcher may need to be creative and resourceful in gaining access to the information required- In terms o: the pursuit ot other objectives in later stages of the research., exp.orniory research is doable-edged. On the one hand, it muy help to establish rapport with individuals or groups being studied and thus smooth rhe way fur later stages of I hr project; on the other hand, it has the potential danger of raising su1piciuii and developing resistance. Therefore the management of exploratory research cannot be taken lightly, as it may have long-term consequences for the prcjec:. In rhe context of his advocacy of symbolic intcraeiiumsm, Biumer ea^e exploratory research a more substantial role. He believed this was nectss.iiy to cour.ier the eommon tendency co niflvr straight into research without an adequate u.iderstsmcin^ of the secret ofsncia 1 life betr.g investigated. He saw the exploratory phase us being necessary to sharpen the focus of the research; not as an optional cura, bur as an essential part of any project-On rhe one hand, i: .s -.he.way by which n researc.a jch:lar nan rorsr. a close aid coiMpre-ieni.ve acqaaaitance with i sphere of social life that is unfamiliar and iiiaic: uakntwn ro lijni |sfc], Ou -he artier ha nil, i; is a means orrfcvelupir.tjtrid sharper: inc. 74 Ressarch Questions and Ch.ecrJves Reietrrcli Objectives T5 his inouary so that hi; problem, his direction! of Mfi.r,\ anslytsa. ttbtOOa* •and iaierprnaiioiis arise aul of, a ad Ktaj:n &»MXX& m, the empirical life una I study. J-yplpraiioa is by definition a flexible -prMávtc in winch the scholar sáJts from one to another line of inquiry, iúo?3 new ponts cf cbscrva.-cm as his study nrogresses, moves in new directions previously ufithought cf, and etacges his reccg-"niciori of what arc iW'i'lEI data £s 1* acquirer more infrnnat.on :nd bener under-staeoing. In ihfic respects, exploratory study stands in ccntrasr n> the prescribed and ciríunvicribed procedure demanded by current tcirnr.fic protocol. Tni flexibility or exploratory pioecdure does not rutin that there is na dirccricn to the -.nquu-v; « means that die toccs U uii&inaUy bioad but beromcs. pnafcfesiivily sharpened « the iucjiii-y prucceds. The purpose of cvpl-raiorv invem&llion is :o move toward a Clearer uuderstancine, ol how one's pro tle ra is to be posed, le- learn what arc appropriate data, to devc'.nji ideas ot what art significant line* of relation, and c« evolve out's conceptual wis in ihe %bl of what one is bannfig arret lbe area 01 life [Blumer l?ť.^. id! Exploratory research řhouid provide as detailed and accurate a picture of the phet.oincnou as is necessary to enable the roearchrr fee! a; home and to he ab:e to speak, about the research problem with some confidence. The picture ptov.de« the tchrlaf with a secure br.-r:r.f to lha? he knows in*: the ^■jesriniis be nski pi tie empiru aj afra *:e meaningful and relevant to it, thai the řiir..Mr:ii In pr.se».not judicial, thai tiie kinds o) data he serks are dj-mftcatf in terrr.s of ihe empirical world, and that the leans he follows ate faithful to its nature, (lilunei l')t&: *3t) lilumcT has icfl us in no doubt about hftW essential exploratory research is to the development n: a ]>ood research design. Dtístription Descriptive research seeks to present an accurate account at some phenomenon, the distribution of characteristics in some population, the patterns ol lelationships in tome social context, at a particular ume, or the changes ir. Ůityt characteristics over tirat (Bulinci \<>M: 6o).Tnt« descriptive accounts can be expressed in words or numbers. They can include the characteristics of a social group or a demographic category, the stages or sequences m social processes, pr patterns in social relationships. They mac in volt: :.-.c development cf sets of categories or types. In practice, the boundary1 ber.veer. expiorstor/ and descriptive research :s blurred. Descriptive research is more ii;4Moui an;! ts usually narrower in its focus « should be directed by clearly stated research 4tteak»n»; However, bota types of research require the use of concepts and they will at structured by a: least srtiiK theoretical aJsumpticms. ExpiinnTiun and \fnfas$é$8nJL ExpiatiEiory research seeks jo account tor patterns i:t observed sociai phenomena, attitudes, behaviour, sociai rcktionships, social processes or social structure* t.bulmrr 13S6: 66-7i. Explsnatiop is making intelligible the events cr re parities that have been observed and whici cannot be accounted for by exkEÍOg tbeo-ics; explanation iliminates pmzks. To explain some phznoocQor. is to gis«c an account of why it behaves in a particular wav or why particular regularities octu:: Explanations provide intellectual satisfaction; they make die obscure plain to sec. Inii is true of both semantic ami scientific explanation. Semantic explanation is concerned with the meanings of words and phrases, while scientific explanation seeks the causes for die occurrence of a particular event or regularity. However, making something mtel.igib-e is not jest a subjective niatter. 'liters is a diffircr.iie bcrween hsi/iKg An ewpJ.ujaT.on and w®Rg it- In th? case vt semarKie explar.ation, wc do nor have one Tinirss ,ind natil wr sec ir. nut :ri thicist cf «tcr,tiric CTphrinrir-r. cither tn: h,ivinj;ar the seeing m.-iy accur v.-i:hiciut the '«ther Tnat an explanation is often resisted when it is first differed is a conirniir.pl.ice of tits history or «cicntc - men |jic- hive it, but do nut sei: it. Tl.e reverse is Lharattcnstit «>: ihe so-t ol expl.in.itiries ntLtirrinj; in ir.yths., p.ir.tnoia, the pecult 'stienecs1, ar.c the like----They ptuvrde a certain .iilellci.tu;,! iiaiistaelifli'Hi I'ut il \i one uilwarriw.i.:! the actual state *; and understanding that has lx;en discussed by writers such as Taylor []9M> ,ir.d vo;i Wright 11971) ;nid. ,(ih%eei explanation and «J*n« explanation. Bxpuihtiicn* ideuiily Causes ul events or rc-galariiies, the factors or mechanisms that produced them, .md understanding ii provideid by the h-asons or accounts social actors i;ive for their actions. The lanet is also assotiated with the niMning of an event or activity in a particular social context, either that given by soaai actors or the meaning that researchers derive from socia. actors'accounts. Rxptaiuithm arc produced by researchers who look at 1 [1.1 = 1 o -l-imi. Ir. .ii dir 'outside', while uKderstandsna i-. I..; sed on ,u, 'inside' \ irw in which researchers isr-asp the subjeetive consciousness, the inte-rpretaticins, of 50cial actor; involved in the tonuuet (tjiddens 1^76: 'I be distinction between crkldtcn (to explain) and vcraeben [to underscsjidi has a long history in German scho!a-shi;>. Wh.ile some writers (e.g. Winch I95x) have argued that caws.il explanation is appropriate in tae natural Sciences *u;i reason explumttion is appropriate m the hmiin:i or socia! scie.it.ces, Otluer wnteri have K&iotl cither mat both.can be used in the social itietitt-s (e.g. Ifabexmas 1972), cr that charactertiing the r«vo fields of science as being exclasivcly conccrnai with Only one of these is inappropriate (e.g. (Hidden., 197a). The position adopti-C here is that both Explanation and tirirfets&iKdikgi&e appropriate objectives in th" social scieares. but thac they produce radier different kinds of intelligibility. Hence, various strategies to achieve explanation or under standing, based on different assumptions and die mc of ibfierent lopes or enquay, have been (6 Hazard; Qucsťais and GfcjccaJves advocated ir. both the natural ana social sciences. These strategics look in different places and ir. terms of different, factors cr mechanisms for answers to their puzzles. In chapter 4, :hree of these explanatory strategies {the inducttve, deductive anil reiToduciwú), and one that is usee! to achieve understanding (the (tbdtutive), will be outlined and compared in terms oftneir relevance to the design a.no conduct of social research, For the credit, a hne: introduction to them will have to suffice. Ir. Lie tr.ductivc research strategy, explanation ii achieved by locating a particular paitcrn within a known and more general pcttem or network of relationships (Kaplan lg64: 293, 333;. The growth of knowledge is achieved by indefinitely filling in and extending the paiterns Tliis form of explanation ii also known as structural cxpkmanon, nsine network or functional theories. Network explanations have three main forms: they refer to broader p.inermof which the case in point can be seen as a sprcitic case thry identify developmental sequences in social rclcuor.sbips; or they specify crrtatii patterns in the way interaction occurs. Functional explanations locate events or patterns of behaviour unii relationships within a larger «r>cial system; the phenomenon bcinf; investigated is explained in terns of die functions i: performs for :he larger system {Neuman 195)7: 53-S). In lite deductive research strategy, explanation is achieved S constructing a deductive argument to which the phenomenon to be explained is the conclusion. The premises ol rir argument will be cither well-established abstract propositions, or hypothetical propositions that are to he tested. While each proposition in the argument nay consist of nothing more than a statement of a rclaiionshir: between two concept;., a sc. of such propositions provides the expliruw.m ;:y linking the lower-level concepts, associated with the described pattern, to more abstract theoretical concrpts. 'I he /rrroi.'urfri'e research strategy seeks to explain a pattern by locating cht causal mechanism that produces it. i'hc pattern of ass.Hriaticm between twe, concepts is usually viewed as entailing a direction of influence, and the existence bf a causal mechanism provides the tXpktndtiaH for the influence. For example, the explanation of the association is?iwee:i religion and occupation, wruch was ot concern to Weber [jySS) as part of his account of die rise of capitalism in Europe, can be explained by tnc mechanism of the mcr.nir.i, particular social actors gave to work.* This type o: causal explanation should not be confused with more common forms which, while not fitting ncady into any of'these three research strategie-, comes closest to the AeiktcUvr. strategy- The classical conception involves a cause and an effect, and Is based on four principles; • temporal order in which the cause must precede the effect; • association which requires that the two events occur together; e the elimination of alternatives in order to be able to claim that the effect was due to the speciiied cause ano not scrncthins else; and / -' V.'n:lc VC:b;r dii oor frame ho rxtihr.irior. ;n t.r.s way. It tvpossin!; ec recaintrji.: h 5cue..=rii witb toe rucm-iia;; of work is £ causal ncthiniso. Tais example wJ. ae taken up 3gaia liiec. I fteseon^i 0, making sense of the causal relationship in terms of broader theoretical idea.. as; emotions. t* Another version involves not jus; one cause bar a sequence or chain of cven-<, ^ vsriahl-s which cumulatively produce :ht effect. While it is normally cxpivsScj diagrammatic form and/or in statisLtctii modets rather than a* a deductive men:, this '.'tew of causal explanation does much the fame as a deductive argument: they both involve connected sequences oi relationships betve-m concepts, which together produce tri expt&nation. The fourth research strategy is dillerer.t from the other three ir, that it is the only one that specifically addresses the objective of u'tdsrsia'tdin^. AbJuclwn is tht logic of enquiry in which enc researcher, at least initially, lakes on the rule of learner and accks to be educated by rhc people bcinj." studied. The initial task is to Icatn about their form of life mid the way they conceptualize and make sense of it, 1 With this knowledge, as 1111 ingredient, the researcher may ihen proceed to rede-sence lay accounts of the social world in social scientific language. The ubjcctiWj is to both describe and understand the pr'oblerri tit It.ud. Wiirccas afpijtttatiwi is concerned wuh abstract patterns of relationships in deductive m^uniems o: causal sequences, tcndci standing is concerned with die reasons social actors give lor their actions. The focus is. noi so much on die explanation* that the researcher constructs bin on the explanations social actors can offer and which enn be Used by die social fCscuchci to coi-stru. 1 .1 .; 1 .il scientific account of their activity. treAxtion Tite objective of iiralktton in research is 10 make claims about what should happen, under certain conditions. Fwdirlttw needs to be distinguished iroai prophecy; the latter makes claimi. about what will happen in the iurua w.itle the lorrner makes claims about what will happen if certain laws oj nu-c.-ianisif-i operate under certain conditions (Popper 12.:!). Prediction involves lime only in the sense thar, whenever pa men's i linvs or mechanisms operate undrr tlvc specified condition*, the predicted outcome can lie expected, Therefore, the p'M-sihihry oi prediction is depc-adem on the state or" k no winder at a nnrricula. QTJJfr Prediction can be achieved in two wnysi in terms of well-established patterns of association between concepts Jas in the inducttt-e research strategy); or by shifting the emphasis in a theoretical argument (as in the defective research strategy:- lfl the cisc of established paitsrns, whenever one parr of a relationship is present, it can be expecrea tftac the other part will also lie present. For example, if it lias be;r> consistently eacailislied rhar juvenile delinquents route from broken homes, then locatirm particular juvenile delinquents carl lead to rite prediction [hat they wit! bp found to have come rrom broken homes, or, alternatively, that children from broken homes arc likely to become delinquents, Some writers have argued that cht: logic involved ir. explatistiafs and }iH>d;cliQn is essentially the fame; it is just a matter of where the emphasis is pur cm", what ram he taken as given {Popper l?5.ii, 1951; Kempt:! 196£;>. This claim is basad on the assLinpriori that a set of propositions thai has been used us an explanation-=f an observed pattern can also be used to predict another pauern. For exsuuotc, if & explanation has teen conjtrtKted to explain why the suicide race is lov.r «b * Research Qucstnons enc1 Qbfee&itš Research Ofcjecťfes country ir. which, a particular religion is predominant, and if religion has been shown in the seductive argument to be related to suicide rates (ss Durkhcim claimed to have estaDlislied), tlacr. it is possible to predict that other countries of a similar rtli^ioiis composition will. ba« similar suicide rates [sec trie discussion o: Hoinans's reconstruction of Durkheim's theory or suicide in chapter 4), Writers who have advocated the retrodttcrivt research stratcp/ le.g. Bhaskat. 1979) have arjeaed that prediction » only pnssibie in closed system: erhaps oil-under experimental condition;. As social scientist' have ti> wnrkjn uren systems, it foilows that prediction is net possible is the social sciences. While explanation in terms of causal mechanisms ispossitlc. fliers is n : :>"e rn:' ■ bec.iu:: the conditions under which a mechanism operates can never be fally established. As the natural sciences also operate in open systems, apart frcrr, artificially con trolled experiments, the advocates of tins position also claim that prediction is not possible in llie natural sciences. Change Heseaich that adopts the objective ufcJsange endeavours 10 intervene in the social world ir» br.nt; about pam.il cr major changes, other in conjunction with the research irself, or as a coitsecjijer.ee of :hc research outcomes. Change can oniy hr ,-uhirved wnh confidence if the actions taken are bavtd on thotc that a -,vcll-established exph'unian nr undcrsundvix would su»res:. However, :hc process of iiucrveniitui itself can be used as a leamitiK process. Knciwlccfje M a phenomenon Ciin be developed in a trial and err«r. process a*imcrv-eritinn iscnndueicd in stages. What is learnt from one stage can be used IV decide what action tfl take in the next st.jr4.c- The outcome car: be uwmiunditig ami r*pisn.tu-n> U well as Jung-. In fact, some philosopher of science Icj;. Popper} have argued that this trial and error process is the only way jicteiitme knowledge can be advanced; that all research involve* Che me of 'piecemeal trchnDlogy" rather than gigantic leaps into unknown territory. Nevertheless, it is possible to distinguish between intervention that iswed primarily fur the purpose of advancing knowledge, and intervention that trie-. to change the world; between purely &eiciiuiic concerns or essentially social or political concerns; between basic research and applied research. The "action research* tradition has the joint objectives of increasing knowledge and changing some aspect r>f the world it the same rime. !t diiiers from more conventional research in that the researcher may take the role of facditaror or resource person ro help a group of |aiun:hfn fes«.irtli, as vvi-,1 as iNtfiM-i ttssca'tit u: ai various kinds, . ■ (ittcenled wi I. |>i:l;« v ,ui.. ;> nj;:aH1iiie ď-,.- npiui-ii Mni nitpl: r.ii i t.i toi. m, particular, and with problem -solving and Uccbiitn-rnakitij; in genera!. Kvaliutuvc rr^circb seeks to examine the consequences tit' the adoption tit particular coupes s( action. It sets out to determine whether a particular policy or programme has i, .u.-ved what i: set out to achieve, that is, iht: exteni to which it has been effective in achieving certain policy or programme goals. Uvaluattun rcscamli compares ■what is" with 'what should be* iWeits 1972; 6): 'The purpose or eva'uaiLui research is to measure the effects of ;t program agamsc ilie goals it set out to accomplish as a means ui coutitbaiitig :o subsequent decision making about tiw program and improving future programmittg (Weiss 1972: ^t. Evaluanon research, has many uses: ♦ tn ĽOjtiinue or discontiiiuä n ;:rfigra:nnie; ♦ to im^ruvs .i pr,,.jjr.i[:tine's practices and procedures; ♦ to add or drop specific program mu strategics and iechr.iq-ae%: ■ to tnstinii; iirnilar ptogranimes e.sewhcrc; ■ to allocate resources amongst txHupecntg progtantmet; and ■ ro accept or reject a pfOgramoW approach or theory [Weiss liř72s lô-'l?; Bu.iner j íáj. However. Weiss bas nrg.ied that, in practice, 'evaiuatiĽni is most often called on to Kelp wita decisions about improving programs. Ga'no gts live or die derisions, are relatively rare' [ 1972: 17). iivaluation research seeks answers Ln ques:u->iis possd by rJtcision-.itiake.-S, net acaden-jc*. However, as LevLie (1937; 30-13 hus pointed oat. in providing Research Qjesnons -and Oi^ectŕcs decision-makers with analysed information, the evaluation researcher has to accept that these decisions arc iikeh to be msJe aceorciinii :o poorly speeded personal, political and naticnal gods. In arguing for 'responsive ccn>tTuetivist evaluation". Cuba and Lmcoi.i (198$) have shifted the focus from decision-makers to the claims, concerns and issues of stakeholders. The tarter include any groups vrh ose stales may be placed ifl jeopardy by the evaluation. Their concern is to adept .v; approach mat will make it possible to i^Ke ihc concerns or aJ staneheldcrs into account. There are two type:, o: evaluation rts-rcrcb.: iurv.aihre evaluation, in which buJt-in monitoring or continuous feedback is used dutín}; the implementation or a policy as a basis for helping to improve it: and summstivi evaluation^ which is conducted after a poke) has been implemented to establish its overall effective-ness in achieving the onpirial goals Pawson and TiJiey 1199?! Have identified lour ntain perspectives on evaluation research: the experimental ;£ittipbeii and Stanley lVrja; Cook and Campbell 1979); the fMgmiie [Weiss 1972, 1976; Wc:s> and Bucuvalas 19801; the tuitur-ahsíic (C-uba and tjncoln 19$y;; and the pluralist (Cronbach 1963. Rossi and freeman íySJ). Tlie firs; on the scene in the 1960*. the experimental per speetivc, used classic:!, or cjuasi-cxprrintcntal procedures :n try to establish whether ihangt is the result of the planned intervention. In the wake of disappointing results Iron this first phase, ihe prapttatic pcrspccttse become less ambitious and advocated the careful use of any kind 0: sound research. The naturalistu- perspective took a diitcrcnt tarn and saw evaluation as a matter pf luy.oti.uioľ. between stakeholders with different mtapietalioas (constructions) of a progi amine, llic pluraltits called foi greater depth and breadth in programme evaluation by c.xatnitmi,; the way pri.grammes arc conceptualized, dealing with both institutional and individual diagnoses ui the problém and fccusuig on outcome effectiveness. Paw son and liilcy have added a fifth pcripccrivc, rtaltttiC evaluation, based an scientific realism l«e chapter 4), fcr which rhcy claim siipcrinnty over the other perspectives. The* have provided eight rules for the conduct of evaluation reseatch. Commonly used tools in both H'jlujuon research anc :n:pait assessment arc needs analysis and uosi-beiiefi! analysis, frowsier, i: a because of the deficiencies in cost-benefit analys.s, cue to its narrow economic focu*. that the development of both sncial impact assessment and environmental impact assessment has occurred. Impact assessment Impt-ci ass;ss.Tier.t TA> has been defined as 'tbcpr-nceis oi ide.'} t l/yr rtg 1 be fuiu re cr. nseq uenees of a current or proposed jcrro.-:' i?.r r kc r 1'.' 7. 2]. In the case of socipl impact assrssmenr i.SlAi, tnese consequences are related to 'indii-iduaU, Cirgsiuzsiiitxi, imtihtíitins and fóttéiy as *i ueole' ,'fiecker iyS'7: Mi), ťpflówing rite definition of SIA La the United States hy the Inier-rtigr.niva:ional Committee on Gaidelir.cs sad Pri.-.cip'.es 111*94;. fturdjc sod Variclay have included cultural as well as social impacts. They include / ail social and cultural í ense emeries haanan pc.pu.4f.otii or an* ja3:i; aai private actions licit alter die way in wa.cn pccplc live, rvi;, reiace ;e or.e another. ßeserj-cfc organise to meet chci; íiccíIí, and girricre] ly cope as .nemhers Ohsociety. Culm > impners i:no.v~ chances ro norms, values, and Delist; of individual r.int t-iird rationalise their oagnirior. bí themselves and dieir society, (Burdge siid Vand WS; II- ' y Recker Limited his definition of SIA to future consequences of pr.eseiu proposed actions. The assessment of the consequence.-, of past actions is scn^. part oi ■ ihiadon research. However, because of the particular coaccwS the practice oi SIA, it is possible to work with a broader definition. Hcmce SLJ,c- 1 be concerned witl a sessing or predicting the demographic, socio-cconoasS insti-.jrional, community and psychological irnpacts of resource ilcvclopniZa and large-scale construction projects, as well as social or economic P'\Im?5 ^ programmes. The tasks of As are to: t assess and predict potential impacts; 1 mitigate and monitor these impacts; and • audit and analyse die impacts 0} past actions. Sc-riat impact assessment auempts tu eiimplement the study ol naural. at biophysical, r:if:nniineiit \ impacts with iiilnrmatiou on the vmial and sociü-i'iyirtftiaic irr facts which may he associated wah a new proiect, pola;y or projy.%nmc. The« it[ is, .;ir .ilter.iliue-. n living c»ndai:>r.s, im Imli- «l,u:i':v. in pv,-. I.i.|.i,;. I 9*6 physi'iloi;ie.d UctWi, eommunity pmcnsi-s, and changes in the prodaciien, distnba-uwi ae-d eotisumpiion «1 goads Jini »et mce». JBtllmftf 08$i 146j I;or example, a niaj«ir road construction scheme may lead to population inove> menrs, the •:.iy,iin-iii.niou of social ooituiiuniues, psychological stress and 1l1.tr.gf! inpropeny values. Siinila: ly, a new social welfare policy may lead to d:sa..v..magci anion;; groups thai ir was supposed to irenclit. SIA will endravnur to ide::nly tkt ran^c and extent of such impacts; 11 can be used tea trade off the benefits m the project te.g. reduced traffic congestion and accidents! against social tt«3». An Important aspect ol iuciii! imp&ci aaeisrttent is the relatiwe gains ar.d Imssea that particular groups iuaconimnr.ity or society arc likely 10 experience as the rc>a:t of a cons: ructton project. Some form ofcompensation for such tosses mii;lit thertix built incb the costing of the project ls.ee a.s:i Finsterbuscii .VH.s, J yJi5). In many ways, SIA has grown out of the related and increasingly itgnifJO. field of envit'.!nmsntitl itnpsct assessment (EIA). While the Inters primary coft* earn is with the natural and biophysical impacts of major physical prc|scts. k a now generally accepted tlia: lilA ;-.nd SIA are complementary and tliat the la^ef must accompany the former. A narrow definition of HA :s rhe assessment of the impact of a planned activity on the environment. However, tt can be conceived more widely as a process fot identifying, predicting and evaluating the biogeographtcal, socic-econcnuc arw human heal to ano welfare consequences 01 iiiipleiasniin^ particular acta*10* rW'adiern 1 r)8S;. The ultimcte purpose is ro indicate to decision-makers :;ie JUeb' consequences of their action, and, therefore, to improve the quality fif decisions iWadnem 1?SS; Orrolsiin &nd Shepherd 1935). ňeiřtuc.4 §uasüahä and Oójectiveí The positive influences of EIÄ are: • withdrawal o: unsound proj-ct.-: • legitimation of sound probers; • selection of improved .ocatioris; • reformulation of plans, for example land use; • redefinition of goals and responsibilities of project proponents Ivan de Gründen 1994t 12-1 is; Ortolano and ehepheid 1SV5: S-9). * Relationships among Research Objectives The four research ch]ectiv-i ai cxpliyrjiion, Jcscriptiun, explanation and prediction can occur as a sequence in terms of boih the Stages and the increasing complexity of research. Exploration usually precedes description, and dvjcripti.r. is necessary before .•xplen.itton of prediction can be attempted, hatploraiior, may be necessary In provide clues about the paiicrr.s that need 10 be described m a p.irutiilar phenomenon. The Nequrr.ce. 'mr.tirmmz wilh i\k description ol patterns, and iü.lowed by an exp!anaiiun of why they occur, central in any form o: social research. Description of -Ahnt it happynuigiqa&io questions o: pu/ known as the 'patte.-r.' model of explanation and is characteristic ol the hiditc'.ii'c research strategy (m be discussed in chapter 4i. There are a variety of views on the relationship between explanation and prediction. It is possible io make prtď.aiMH without having an explanation of a phenomenon. This kind of predittiun reules on well-established gerterraliutioiis about pa: lern» of relationships betwern concepts, somriimes rcfrrrrc to as statistical law*. It is s common belief that the natural sciences aim to establish mwwwl faiitf about the regularities in die natural world, laws that are claimed to hold at all times and in all places. Whet :i-r it is pois.hle to establish ->uch laws in the social sciences :s a matter of considerable dilute. What might be claimed i> taut tie social sciences can establish patterns or regularities, but that these arc limited :n time ar.d space. The conccpr of statistical law :±:c.-j ro such regularities. While »yrac philosophers have arr:ued that üieje patterns provi.de a basis for explain: I ion, others have argued that it is necessary :o find the mechanism tliat produces s.ich patterns before explanation can be achieved istc chapter --). i low-ever, the description of patterns or rclatioiivli ps between concepts can be usee; fc-r prndic-tion. The objectives of eifoluation and impact assessment share much- in commoru Thcy, together with tr.icn.vtv.icn, ccnstittrte the main fields of applied research. Ai we have seen, n major d.s~inguisltmg feature of applied research is thai it has a sponsor and/or ckcr.t. Its goals arc either se; by the Sponsor, or arc the outcome of negotiation between the sponsor and researcher, arid its outcomes have to address the concerns of the client. While it may be possible to attempt evaluation and impact assessment from an acbcoieticsi point of view, by builcinj; or. only a descriptive research base and side-stepping the objectives of explanation o: understanding; sophisticated awjiuation and impact assessment, need to use existing theories. If relevant theories are not available, they will need to be developed. Because applied research is normally clone within severe tune and resource constraints, there is pressure to take short cuts to avoid ihese essential components. Good applied research has to draw on well-csiablished theor.es, because, after all, there is frothing as practical as a ;$ood theory. It is unlikely that the whole gambit of objectives can be or need to be tackled in most research protects, and certainly not within the limitations of post-uraduute research. Previous research may l.c.ve achieved some objectives that can be used as a background in a pirncnlnf research project. l:or example, if good descriptive research has already been done in the iield, it nay be possible to begi*; with an explanatory objective, or if wcl.'-establishi-d and relevant theories arc available, it may be possible to engage directly in the objectives of tlian^e, rvatua-utin Of impact assessment. Uut, to repeat an earlier comment, withoiu in iiUq i.iie ilc»cr:p::ve base, it is not possible to begin to pursue the other research objectives. Research Objectives and Questions Each of the eight research objectives is related to a particular type of research question. If wc take some imaginary social process as an example, the three type* of tesrarch questions would be associated with the eight research objectives uc as follows. Expianttton What Might be happening? What people arc invoked? In whai w .-.,■! Description What is happening? What people art involved? In what way? UrzdirsUir.dm% Why rs it happening? Explanation Why is it happening: Prediction What is likely to happen? CJiuige How can it be made ro be different? Evaluation What lias happ.er.ed? Why did it ,-iappenr Asisdsnient Wliaf have been, or arc lilscly tq lie, iis Individual, social and environmental conseqaencts? Why have these consequents oeiritfredr The oajecuves of undemanding and cxpiamtion? and. to a lesser extent, vjaluaiion -and impact assessment;, arc the only ones ihat require \vhy'-type ' Whiir n»ia<;,-jra«iii.^-ird eXfitOHr^Bn email rl-.e fume sinJ u(rtaeanth que=tiei:i, :t i» rtccrssirv :<\ toeir. sc-juae us ton-arch clvnitivo; as they ans.vc: 'iv.ij' qtwsttonj .n Sttfc/int v.»),s. 84 Pcsccrch Questions end Objectives questions. Change :s the cniy objective that require!; 'how'-rype questions. All ice other objectives have question* begianinc with 'wh3t\ or their questions tan he transposed into this form. They are. therefore, cither descriptive in nature, or involve ccr.tpariscns hcr.vecn ;ituati:ns in tr.c present, between a present and a past situation, or between a present situation and a desirec future. To avoid the conlnsa.in tli.it can result from other question wording, for example the pursuit nr the obje-cinrs of (iejcriplhn or ext'knutivn wlh questions that commence with 'how', this three-category ciassifkatiua of qttcsriens should be followed. Conclusion By way of summary, let me review the key points that emerRC from this discussion of the role ot research questions in social resecrch- » All research projects ate built on the foundation of rrs&irch questions, • Researen questions define the " ktufe arid scope of a research project. • Research questions can be grouped into three main rypes, 'what', "why* and 4howv questions. • The three type* of questions form a sequence for the research process; 'w-hat' questions followed by ■why' questions followed by "how* questions. « Many reset::!', nhicctivcs require 'whit' qufsnons. It is only the aims of under-ftanthag and expisnatian, one possibly evaluation and unpad atscuirxn:, that require 'why' question-;, and the aim of change that requires "how* questions. • "Die importance of answering 'what' question* should n.'i underestimated. • The developmental nature o: a rcscatch design >ltould not be used as an excuse to avoid the effort required to formulaic appropriate research questions. • While the process ol developing a set of research questions caa be the most challenging part of any research prop;;:, techniques are availab.c to assisr the process. • Research questions are what the research is designed :o answer, not the questions asked of respondents or paniciparws, • The aim of the literature review is to indicate what the state of knowledge is with respect to each research question, or group of questions. ■ 1 lypotbe-scs arc our best )>csses ai answering 'why* and, puvJb'y, 'how' questions. • If required, hypotheses should be derived frcrr. the literature renew, particularly from theory or research results. On rare occasions, a theory may have to be generated. • In some research, hypotheses may emerge, and be tested, in the course uf the data collection and analysis. As an aid to the conception, clarification ar.j classification of research ques-tioits, it is useful to chink about a research project in terms ot its objectives, ffie"* olvieciivesare nor a list of the activities tac research.-. gGing to carry out. Rather, tlif.y fjafl he either the analytical or ihx practice aims of a projecr. I Strategies for Answering Research Questions ! ir'crvy iNifw'ry must s/urj iwtvivhrtr. (KaisÍMI 1964,1 Introduction 1 laying established a .set of research questions, the next calk is to devise ways to answer tiietn. 'I he approach taken id answering research questions depends on the type of question. Answering a 'what' question is usually easier than rtrwWrcrirtg a Sfiřby" or 'how' question. 'What' questions tan he ..call with by making appro* priatc observations or measurements, i.e. collecting appropriate data, and then producing descriptions bused on them, i lovvevrr, tins process is not .is simple as it sounds; descriptions of what we believe we have observed may not be, perhaps cannot be. pure descriptions. Hie observer, as an active participant in thu precesi, has to make many decisions before ,i description tan Ive produced, and cat nor *vcitl imposing concepts and categories. Answers ro 'bow' questions require a diifcrcnt kind of description; b po.-> Uk state of affairs Iws to he described and ideas about how to get rhcrc nave to he provided. As we saw ;n chapter 3, 'how' questions usually require answers to rrlated 'what' and 'why' questions, either in the research being undertaken, or in previous research. Unless a good intlersraiid.i^ of the nature of the phenomenon being investigated has already been achieved, and why ír behaves the -way it ikies, it is difficult, undesirable aud even dangerous to begin to propose any form of intervention. [ lowcvc-i, the monitoring of limited int,er?ca">-tions in 'safe' situations (i.e. ones rha: will not have ethically undesirahe or socially unacceptable consequences) is one way of discovering answers to 'why' and 'how' questions. Action research is iudi a learning process. 1 will come back to the ways of answering 'what and 'how' questions later in the chapter, in the rneamtime. I want to conccntrart on Itnty to answer 'why' questions. The miinprabkrriinanswertng'wiiy'quesrioiis :s where to look for thesnswitR 1 low wo deal with this w:I] decern i uc where lb- research process bef$fi&and how it