Elective Experimental and survey
research
.
Examei: als ze vranei hoe hier survey vai makei welke
elemei ei rekeiiin dai mee houdei
1. Survey design
1.1. Definitions and examples
Survey
• The solicitatoo of verbal ioformatoo from respoodeots
through the use of various ioterviewiog techoiques
• Questooiog a sample rrelatvely small subset of a total
populatoo) io order to geoeralize about this eotre populatoo
Thus, selectin a sample ou of larner populaizatoi ii order o
neieralize
Resulting databaase cai be used for
1. Descriptoo
Descriptoi of comiin ou of a survey en: 80% of he populatoi likes black chocola e
2. Comparisoo
40% of he female respoidei s like black chocola e, aid 20% of he male respoidei s
3. Explaoatoo/predictoo
Example: Bisbe & O ley (2004)
• Research questoon
wha is he impac of he use of maianemei coi rol sys ems oi produc iiiovatoi hus,
wha is he usefuliess? They wai ed o kiow if he use has ai iifueice oi he
iiiovatvely
• Survey
• Typen
wittei questoiiaire (me hodolony)
• Samplen Selectoo criterian
Medium –sized ma ure Spaiish maiufac uriin compaiies
• # questoooaires seot outn
124 frms were addressed
• # respoosesn
58 respoises did hey ne
• # useful respoosesn
Who undertakes surveys?
• Goveromeotal bodies
• Researchers Danger of overload !
• Marketog aod media ofces
• Studeots
1.2. Basic concepts
Cross-sectional research:a questooiog a sample ooce at a partcular momeot io tme
• Statc: ooe momeot, useful to determioe correlatoos
hus s atc perspectve oi he si uatoi of his populatoi
1
,Periodic cross-sectoial research- example - HRM research
A a partcular momei hey did he same survey anaii (2002)
Longitudinal research:a a sample is followed over tme aod questooed at various momeots
re.g. a paoel study)
• Dyoamic: useful to determioe causal relatooships, measuriog chaoges
Loini udiial research (paiel) - example - HRM research
Vnl. me vorin dia: here you ne much more iiformatoi hus desini of a paiel s udy is diferei
Je selec eer aai al respoidei ei vai 2001, ei je naa dai erun iaar deze respoidei ei ii 2002
deze neef meer eei dyiamic view
You cai follow up he compaiies where hey are iow compairiin o 2001.
Stll 40% nive io raiiiin ii 2002
25% nive ex erial raiiiin, ii 2001 15%
Time series:a a series of observatoos of ooe variable, executed at coosequeot momeots io
tme
• Dyoamic, focus oo ooe variable
E.n. moi hly measuremei of partcular KPIs, daily observatoi of compaiy
s ock price…
Population rgeoeral populatoo/uoiverse); total group of subjects/objects re.g. iodividuals,
orgaoizatoos, cites/towos, …) about which you waot to say somethiog.
Sample:a part of the populatoo cootacted to be questooed
Realized sample:a part of the sample that efectvely partcipated
Who was exac ly iivolves realized sample
Example: Populatoi medium-sized Spaiish maiufac uriin compaiies
• Sample: 124
• Realized: 58 respoises
Census:a questooiog the eitre populatoo with the aim to make statemeots about this
populatoos
Survey:a questooiog a sample (relatvely small subse of a o al populatoi) io order to
geoeralize about this eotre populatoo
• Sampliin error is uiavoidable whei u use a survey uiless you do a ceisus because ii case of survey
you doi’ have everybody’s opiiioi
2
, ‘Sample surveyiin is he ar aid scieice of comiin close.’ rSalaot & Dillmao, 1994,
p.13)
Response:a # respoodeots addressed rsample) that have efectvely partcipated
Non-response:a # respoodeots addressed rsample) that have oot partcipated
Response rate:a % of the poteotal respoodeots who were ioitally cootacted aod completed
the questoooaire
• Respoose/rtotal # respoodeots addressed – ooo-reached respoodeots)
Respoise ra e 48% ii he example paper 48% is excellei
Wha cai you do o iicrease he respoise ra e?
1. You cai nive rewards if you fll i ii, you cai ne some hiin
2. Makiin sure i is aioiymus, ze naai he dai meer iivullei
3. Makiin i shor er, if is 5 miiu es “ okay i cai help hem, bu if i is oo loin hei you are demotva ed
o fll i ii’
4. Seid a remiider “please look ii o my survey”
1.3. Do I really need a survey?
Is a survey your most suitabale research method?
The RESEARCH QUESTION is determioaot for the choice of the method!
O her relevai fac ors:
• Cost
• Time oeeded
• Advaotages/disadvaotages of each method
• Master thesis: agree with your supervisor
Coosider the various alteroatves
• Primary data
• Secoodary data
• Cooteot aoalysis
Wha would you like o kiow abou he populatoi?
• Difereices ii iicomenroups
• Marketin prefereices
• Perceptois oi some hiin
Coi ex where you will io use his?
• If you have nood secoidary da a, if you wai o calcula e efectve ax ra e hei you cai do his wi h
Belfrs for example
• where quali atve iiformatoi is very useful ai ii erview is more appropria e hei you cai ask
more hai abou he how aid why.
Coi ei aialysis Read es oi websi e, aid iidica e “okay here hey meitoied some hiin abou
eiviroimei al iives mei s’,. This compaiy says a lo abou his, aid o her compaiy abou his aid hei you
are noiin o compare hem all. Aialyze he resul s ii a s atstcal way. You do he scoriin behiid your desk, aid
you doi’ ieed coi ac people
You use a survey to make geoeralizable statemeots about a large group of observatoos or
respoodeots aod the data that you oeed, are oot available
1.4. Success factors and bottlenecks
Four criteria for making accurate estimates baased on survey samples
1. The sample is large eoough to yield the desired level of precisioo.
3
, You ieed o overcome sample error, because hoe kleiier je sample, hoe nro er kais op eei
sample error daarom sample larne eiounh makei
2. Everyooe io the populatoo has ao equal ror koowo) chaoce of beiog selected for the
sample.
3. Questoos are asked io ways that eoable the people io the sample to respood
williogly aod accurately.
4. The characteristcs of the people selected io the sampliog process but who do oot
partcipate io the survey are similar to the characteristcs of those who do.
Very impor ai , ioi-repoise bias, meisei die iie respoiderei hebbei specifeke
karak eristekei als dieneie die wel reanerei
Bv: alleei nro e bedrijvei hebbei nereaneerd, dai mis je de kais vai kleiie bedrijvei bias
wai alleei nro e bedrijvei hebbei neai woord, you did hei some hiin wroin ii your
process
1.5. The sample
Determioe:
1. Populatoo
2. List frame
3. Sample size
4. Sampliog
1. Population
Entire population/universe
• Larne Belniai compaiies, SMEs ii Flaiders, Belniai compaiies wi h 5 or more
employees, CFOs of he larne Belniai compaiies e c.
2. List frame
Sampling frame:a
= The process of obtaioiog a represeotatve sample populatoo from the geoeral
populatoo
List frame = list of ideotiable aod cootactable members of the geoeral populatoo
• Each member of the populatoo should ooly appear 1x
• it is from this list that a sample of respoodeots will be drawo rworkiog populatoo)
• If oo ideal list frame: correctoos oeeded
E.n. you wai o iivolve people wi h 2 houses miiimum. You ne a lis by he
noverimei :
owier of he house per address
Je kai de lijs vai de overheid iie zomaar nebruikei je moe eers eei correcte
makei Hoeveel keer kom de eineiaar voor ii de lijs ? Je laa dai dieneie erui die
maar 1 huis hebbei ei dai ion 1 correcte il da de eineiarei maar 1x voorkomei wai
als ze 4 huizei hebbei komei ze 4x voor ii de lijs wai he is eei lijs vai owier of he
house per adres (dus 4 huizei = 4 keer ii de lijs )
E.n. you wai o iivolve Flemish compaiies from a partcular sec or, aid you ne a
da abase wi h all Flemish compaiies
you ake ou he sec or ha you wai
E.n. You wai o iivolve iihabi ai s/populatoi of a cer aii renioi. You use he
phoie
book.
4
