The Sources of Data

12
Thedatausedinempiricalanalysismaybecollectedbyagovernmentalagency(e.g.,theDepartmentofCo
mmerce),aninternationalagency(e.g.,theInternationalMonetaryFund(IMF)ortheWorldBank),apriv
ateorganiza-
tion(e.g.,theStandard&Poor’sCorporation),oranindividual.Literally,therearethousandsofsuchagen
ciescollectingdataforonepurposeoranother.
The Internet
TheInternethasliterallyrevolutionizeddatagathering.If
youjust“surfthenet”withakeyword(e.g.,exchangerates),youwillbeswampedwithallkindsofdatasour
ces.In
Appendix E
weprovidesomeof thefrequentlyvisitedwebsitesthatprovideeconomicandfinancialdataof
allsorts.Mostofthedatacanbedownloadedwithoutmuchcost.Youmaywanttobookmarkthevariouswe
bsitesthatmightprovideyouwithusefuleconomicdata.Thedatacollectedbyvariousagenciesmaybe
experimental
or
nonex-perimental.
Inexperimentaldata,oftencollectedinthenaturalsciences,theinvestigatormaywanttocollectdatawhile
holdingcertainfactorscon-
stantinordertoassesstheimpactofsomefactorsonagivenphenomenon.Forinstance,inassessingtheimp
actofobesityonbloodpressure,there-
searcherwouldwanttocollectdatawhileholdingconstanttheeating,smok-
ing,anddrinkinghabitsofthepeopleinordertominimizetheinfluenceof
thesevariablesonbloodpressure.Inthesocialsciences,thedatathatonegenerallyencountersarenonex-
perimentalinnature,thatis,notsubjecttothecontroloftheresearcher.
13
Forexample,thedataonGNP,unemployment,stockprices,etc.,arenotdirectlyunderthecontrolofthein
vestigator.Asweshallsee,thislackofcontroloftencreatesspecialproblemsfortheresearcherinpinningd
owntheexactcauseorcausesaffectingaparticularsituation.Forexample,isitthemoneysupplythatdeter
minesthe(nominal)GDPorisittheotherwayround?
The Accuracy of Data
14
Althoughplentyofdataareavailableforeconomicresearch,thequalityof
thedataisoftennotthatgood.Thereareseveralreasonsforthat.First,asnoted,mostsocialsciencedataaren
onexperimentalinnature.Therefore,thereisthepossibilityofobservationalerrors,eitherofomissionorc
om-mission.Second,eveninexperimentallycollecteddataerrorsofmeasure-
mentarisefromapproximationsandroundoffs.Third,inquestionnaire-
typesurveys,theproblemofnonresponsecanbeserious;aresearcherisluckyto
geta40percentresponsetoaquestionnaire.Analysisbasedonsuchpartialresponsemaynottrulyreflectth
ebehaviorofthe60percentwhodidnotrespond,therebyleadingtowhatisknownas(sample)
selectivitybias.
Thenthereisthefurtherproblemthatthosewhorespondtothequestionnairemaynotanswerallthequestio
ns,especiallyquestionsoffinanciallysensi-
tivenature,thusleadingtoadditionalselectivitybias.Fourth,thesamplingmethodsusedinobtainingthed
atamayvarysowidelythatitisoftendiffi-
culttocomparetheresultsobtainedfromthevarioussamples.Fifth,eco-
nomicdataaregenerallyavailableatahighlyaggregatelevel.Forexam-
ple,mostmacrodata(e.g.,GNP,employment,inflation,unemployment)areavailablefortheeconomyas
awholeoratthemostforsomebroadgeo-
graphicalregions.Suchhighlyaggregateddatamaynottellusmuchabouttheindividualormicrounitstha
tmaybetheultimateobjectofstudy.Sixth,becauseofconfidentiality,certaindatacanbepublishedonlyin
highlyaggregateform.TheIRS,forexample,isnotallowedbylawtodisclosedataonindividualtaxreturn
s;itcanonlyreleasesomebroadsummarydata.Therefore,ifonewantstofindouthowmuchindividualswi
thacertainlevelofincomespentonhealthcare,onecannotdothatanalysisexceptataveryhighlyaggregate
level.Butsuchmacroanalysisoftenfailstorevealthedynamicsofthebehaviorofthemicrounits.Similarl
y,theDepartmentof
Commerce,whichconductsthecensusofbusinessevery5years,isnotallowedtodiscloseinformationon
production,employment,energycon-
sumption,researchanddevelopmentexpenditure,etc.,atthefirmlevel.Itisthereforedifficulttostudythei
nterfirmdifferencesontheseitems.Becauseofalltheseandmanyotherproblems,
the researcher shouldalways keep in mind that the results of research are only as good as
thequalityofthedata.
Therefore,ifingivensituationsresearchersfindthattheresultsoftheresearchare“unsatisfactory,”thecau
semaybenotthattheyusedthewrongmodelbutthatthequalityofthedatawaspoor.Unfortu-
nately,becauseofthenonexperimentalnatureofthedatausedinmostsocialsciencestudies,researchersv
eryoftenhavenochoicebuttodependontheavailabledata.Buttheyshouldalwayskeepinmindthatthedat
ausedmaynotbethebestandshouldtrynottobetoodogmaticabouttheresultsob-
tainedfromagivenstudy,especiallywhenthequalityofthedataissuspect.
ANote on the Measurement Scales of Variables
15
The variables that we will generally encounter fall into four broad cate-gories:
ratio scale, interval scale, ordinal scale, and nominal scale.
It is im-portant that we understand each.
Ratio Scale
For a variable
X
, taking two values,
X
1
and
X
2
, the ratio
X
1
/
X
2
and the distance (
X
2

X
1
)are meaningful quantities. Also, there is a
natural ordering (ascending or descending) of the values along the scale.Therefore, comparisons
such as
X
2

X
1
or
X
2

X
1
are meaningful. Mosteconomic variables belong to this category. Thus, it is meaningful to
askhow big is this year’s GDP compared with the previous year’s GDP.
Interval Scale
An interval scale variable satisfies the last two propertiesof the ratio scale variable but not the
first. Thus, the distance between twotime periods, say (2000–1995) is meaningful, but not the
ratio of two timeperiods (2000
/
1995).
Ordinal Scale
A variable belongs to this category only if it satisfies thethird property of the ratio scale (i.e.,
natural ordering). Examples are grad-ing systems (A, B, C grades) or income class (upper,
middle, lower). Forthese variables the ordering exists but the distances between the
categoriescannot be quantified. Students of economics will recall the
indifference curves
between two goods, each higher indifference curve indicating higherlevel of utility, but one
cannot quantify by how much one indifference curveis higher than the others.
Nominal Scale
Variables in this category have none of the features of the ratio scale variables. Variables such as
gender (male, female) and mari-tal status (married, unmarried, divorced, separated) simply
denote cate-gories.
Question:
What is the reason why such variables cannot be expressedon the ratio, interval, or ordinal
scales?As we shall see, econometric techniques that may be suitable for ratioscale variables may
not be suitable for nominal scale variables. Therefore, itis important to bear in mind the
distinctions among the four types of mea-surement scales discussed above.

Sign up to vote on this title
UsefulNot useful