RMCT 11 DATA ANALYSIS BTM

8/19/2019 RMCT 11 DATA ANALYSIS BTM

1/52

Research Methods in Computing

and TechnologyCT098-3.5-2

DATA ANALYSIS


2/52

CT098-3.5-2 Research Methods in Co!"t in# $ Techno%o#& 02 Research $ Li terat"re Searches

Topic & Structure of the lesson

• Ana%&sin# '"antitati(e Data – Techni)"es

–

Considerations *or )"antitati(e data ana%&sis


3/52


Learning Outcomes

At the end o* this to!ic+ &o" sho"%d ,e a,%e to

• /!%ain )"antitati(e data and )"antitati(e

ana%&sis techni)"es

• Descri,e the (ario"s considerations *or

)"antitati(e data ana%&sis


4/52


'"antitati(e Data

• Ra *or

• Con(e& %itt%e eanin# to ost !eo!%e

•

N"erica% data• Sho"%d he%! to

– Anser research )"estions

–

Meet research o,1ecti(es• Needs to ,e ana%&sed and inter!reted

)"antitati(e ana%&sis techni)"es


5/52


'"antitati(e Ana%&sis Techni)"es

• /a!%es – Creatin# si!%e ta,%es or dia#ras that sho

*re)"enc& o* occ"rrence

– se statistics to ena,%e co!arisons

– Co!%e/ statistica% ode%%in#


6/52



7/52CT098-3.5-2 Research Methods in Co!"t in# $ Techno%o#& 02 Research $ Li terat"re Searches

Considerations *or )"antitati(e

ana%&sis

• T&!e o* data

• 4orat data i%% ,e in!"t to the ana%&sis

so*tare %a&o"t• I!act o* data codin# on s",se)"ent

ana%&ses

•

Need to ei#h cases• Methods "sed to chec data *or errors



T&!e o* data

Cate#orica%

Descri!ti(e Raned

'"anti*ia,%e

Interna% 6Ratio Data

Contin"o"s 6

DiscreteData



Cate#orica% Data

• cannot ,e eas"red n"erica%%&

• can ,e c%assi*ied into sets 7cate#ories accordin#

to characteristics

• can ,e !%aced in ran order

• can *"rther ,e s",di(ided into descri!ti(e

7noina% data and raned 7ordina% data



Descri!ti(e Data

• Aa noina% data

• # Cate#orisin# Cars

• I!ossi,%e to de*ine cate#or& n"erica%%& 6 ran

it

• Co"nts the no. o* occ"rrences in each cate#or&

• Cate#ories sho"%d ,e "na,i#"o"s 6 discrete

• I* on%& 2 cate#ories a%e 6 *ea%e dichotoo"s data:



Raned Data

• Aa ordina% data

• More !recise *or o* cate#orica% data

•

;no the re%ati(e !osition o* each case• # Ranin# 6 sca%e )"estions



T&!e o* data

Cate#orica%

Descri!ti(e Raned

'"anti*ia,%e

Interna% 6Ratio Data

Contin"o"s 6

DiscreteData



'"anti*ia,%e Data

•



Inter(a% 6 Ratio Data

• Inter(a% data

Yesterday = 20 deg C

Today = 30 deg C

Difference = 10 deg C not 1.5 degrees warmer

• Ratio data

Last year = RM3 m

This year = RM m

Ratio = do!"#e



Contin"o"s 6 Discrete Data

• Contin"o"s data – (a%"es that can tae an& (a%"e

– # #ength of ser$ice% tem&erat!re of f!rnace%

• Discrete data – can ,e eas"red !recise%&

– # no. of c!stomers



= What factors determine the most

appropriate statistical techniques?

= Research o,1ecti(es 6 )"estions 6 !"r!ose o*

&o"r st"d&

= Meas"reent sca%es &o" "sed in &o"r

research instr"ent

= Research desi#n o* &o"r st"dies.



'"ic Re(ie '"estion

= Di**erentiate ,eteen

= Inter(a% and ratio data

= Contin"o"s and discrete data= Descri!ti(e and raned data

= >ro(ide a!!ro!riate e/a!%es to

i%%"strate the di**erences



Data La&o"t

• Data is "s"a%%& stored in atri/ 6 ta,%e

*orat


19/52


Codin#

• A%% data t&!es sho"%d ,e recorded "sin#

n"erica% codes – '"icer "se n"eric e&!ad %ess errors

– S",se)"ent ana%&ses ore strai#ht*orard

• Codin# )"anti*ia,%e data – Act"a% n",ers are "sed

– Recodin# e.#. Sa%ar& entered as

)"anti*ia,%e discrete data then re-coded as

cate#orica% raned data


20/52


Codin#

• Codin# cate#orica% data – Need to desi#n and "se a codin# schee

• enco"ra#ed to "se e/istin# codin# schee. ?h&@ – Sa(es tie

– Nora%%& e%% tested

– A%%os co!arisons ith other 7%ar#er s"r(e&s

– Inc%"de on data co%%ection *or as !re-set

codes 7%iit the n",er o* cate#ories


21/52


Codin#

• Codin# a*ter data co%%ection - necessar& – "nc%ear o* %ie%& res!onses

– Lar#e n",er o* !ossi,%e res!onses in the

codin# schee

• Code,oo de(e%o! codin# schee a*ter

co%%ectin# *ro 50 B00 cases


22/52


Codin#

• Codin# issin# data – sed to indicate h& data are issin#

• Data o!tiona% not re)"ired *ro e(er& !artici!ant

si! #enerated ,& a *i%ter )"estion

• Non-res!onse res!ondent re*"se to anser

• Res!ondent did not no the anser 6 ha(e an

o!inion

• Res!ondent issed a )"estion ,& istae

• Res!ondents anser "nc%ear


23/52


Codin#

– S!ecia% code *or issin# data di**erent codes *or

di**erent reasons *or issin# data

– Not issin# data %ea(in# !art o* a )"estion in a

s"r(e& ,%an indicates an anser


24/52


nterin# data

• ns"re data entered correct%& - IE

• Data ana%&sis so*tare –

a%#oriths that chec the data *or o,(io"serrors

– A%%os researcher to attach %a,e%s to each

(aria,%es and the codes associated ith each

o* the• Red"ces isinter!retation hen ana%&sin# data


25/52


26/52


?ei#htin# cases

B. Ca%c"%ate the H o* the !o!"%ation res!ondin#

*or each strat"

2. sta,%ish hich strat" has the hi#hest H o*

the !o!"%ation res!ondin#

3. Ca%c"%ate ei#ht *or each strat"

hi#hest !ro!ortion !o!"%ation res!ondin# *or

an& strat" 6!ro!ortion o* !o!"%ationres!ondin# in strat" *or hich ca%c"%atin#

ei#ht

. A!!%& the a!!ro!riate ei#ht to each case


27/52


'"ic Re(ie '"estion

Descri,e hat a researcher o"%d need to

consider hen doin# a )"antitati(e ana%&sis


28/52


To sho s!eci*ic (a%"es


29/52


To sho hi#hest and %oest

7cate#orica% and discrete data


30/52



7cate#orica% and discrete data


31/52



7contin"o"s data

Histograms ith e!en class inter!als


32/52



7contin"o"s data

Histograms ith une!en class inter!als


33/52



7contin"o"s data

4re)"enc& !o%on


34/52


To sho trend

Line #ra!h


35/52


To sho !ro!ortions

>ie Chart


36/52


To sho distri,"tion

7Contin"o"s data

4re)"enc& !o%on Gisto#ra


37/52


To sho distri,"tion

7Discrete data

4re)"enc& !o%on Jar chart


38/52


>ositi(e%& 6 Ne#ati(e%& Seed


39/52


S&etrica%%& Seed

7Nora% Distri,"tion


40/52


Jo/ !%ot

Co!arin# (aria,%es sho


41/52



s!eci*ic (a%"es and

interde!endence

Contin#enc& ta,%e 6 cross-ta,"%ation


42/52


Co!arin# (aria,%es co!are

hi#hest and %oest (a%"es

M"%ti!%e ,ar chart


43/52



!ro!ortions

>ercenta#e co!onent ,ar chart


44/52



trends

M"%ti!%e %ine #ra!h


45/52



tota%s

Staced ,ar chart


46/52



!ro!ortions and tota%s

Co!arati(e >ro!ortiona% >ie Charts


47/52



distri,"tion o* (a%"es

M"%ti!%e *re)"enc& !o%ons


48/52




M"%ti!%e ,ar charts


49/52




M"%ti!%e ,o/ !%ots



50/52



re%ationshi! ,eteen cases *or

(aria,%es

Scatter #ra!h


51/52


'"ic Re(ie '"estion

/!%ain hen and h& &o" o"%d "se each o* the *o%%oin#

charts 6 #ra!hs 6 ta,%es hen ana%&sin# )"antitati(e data

– Scatter #ra!h

–

M"%ti!%e *re)"enc& !o%ons – Contin#enc& ta,%e

– Staced ,ar chart

– Gisto#ra


52/52

' $ A

"uestion and #nser Session

RMCT 11 DATA ANALYSIS BTM

Documents