Results Analysis and Discussion Workshop

8/21/2019 Results Analysis and Discussion Workshop

1/52

Results Analysis andDiscussion Workshop

By

Charles Bong Hin Joo


2/52

Results

“Results! Why, man, I hae gotten a lot oresults" I kno# seeral thousand things that#on$t #ork"%

& 'homas A" (dison


3/52

Content o the Results

Do not start the Results section )y descri)ing methodsthat you inadertently omitted rom the *aterials and

*ethods section"

'#o key ingredients o the Results section+

a -ie some kind o oerall description o thee.periments, proiding the “)ig picture%, #ithout

repeating the e.perimental details preiously proided in*aterials and *ethods"

) /resent the data in the past tense"


4/52

(.ample 0 Results

Bong, C"H"J" 12345" Self-cleansing urban drain using sediment

flushing gate based on incipient motion" /h"D 'hesis, 6niersiti 7ains*alaysia"

-ie oeralldescription oe.periments


5/52

Content o the Results

/resent representatie data rather than endlesslyrepetitie data 1you can attach the rest o the data inAppendi."

“'he compulsion to include eerything, leaing nothing out,does not proe that one has unlimited inormation8 itproes that one lacks discrimination% & Aaronson 149::

“'he ool collects acts8 the #ise man selects them%

& Wesley /o#ell 14;;;


6/52

Ho# to Handle


7/52

Ho# to Handle


8/52

7trie or Clarity

'he results should )e short and s#eet, #ithout er)iage"

“I you are out to descri)e the truth, leae elegance to thetailor%

& (instein

Although the Results section is the most important part,it is oten the shortest, particularly i it is preceded )y a#ell0#ritten *aterials and *ethods section and ollo#ed

)y a #ell0#ritten Discussion"

'he Results need to )e clearly and simply stated)ecause it is the Results that constitute the ne#kno#ledge that you are contri)uting to the #orld"


9/52

Aoid Redundancy

'he most common ault is the repetition in #ords o#hat is already apparent to the reader rom e.aminationo the igures and ta)les"

(en #orse is the actual presentation, in the te.t, o allor many o the data sho#n in the ta)les or igures"

Do not )e er)ose in citing igures and ta)les" Do not

say “It is clearly sho#n in 'a)le 4 that nocillin inhi)itedthe gro#th o N. gonorrhoeae"% 7ay “


10/52

'a)les

“A ta)ular presentation o data is oten theheart or, )etter, the )rain, o a scientiicpaper"%

& /eter *organ


11/52

When to 6se 'a)les

As a rule, do not construct a ta)le unless repetitie datamust )e presented"

I you make 1or need to present only a e#determinations, gie the data in te.t"


12/52

(.ample & Repetitie Data

Bong, C"H"J", >au, '">" ? A)" -hani, A" 12345" Hydraulics characteristics o tippingsediment lushing gate" Water Science & Technology , 68144, 259:02@3"


13/52

When to 6se 'a)les

'a)le 4 is aulty )ecause t#o o the columns giestandard conditions, not aria)les and not data"

'he data presented in 'a)le 4 can )e presented in the

te.t itsel, i"e" “Aeration o the gro#th medium #asessential or the gro#th o Streptomyces coelicolor " Atroom temperature 12@C, no gro#th #as eident instationary 1unaerated cultures, #hereas su)stantialgro#th 1D, :; lett units occurred in aeratedcultures"%


14/52

When to 6se 'a)les

In 'a)le 2, the independent aria)le 1temperature looksreasona)le, )ut the dependent aria)le column 1gro#thhas a suspicious num)er o Eeros"

'a)le 2 is useless )ecause all it tells us is that “'he oakseedlings gre# at temperatures )et#een 23 and 53C8no measurea)le gro#th occurred at temperatures )elo#23C or a)oe @3C"


15/52

When to 6se 'a)les

'a)le 5 is not inormatie" All the ta)le tells us is that “S. griseus, S. coelicolor , S. everycolor , and S.rainbownesky gre# under aero)ic conditions, #hereas S.nocolor and S. greenicus reFuired anaero)ic conditions"%

Wheneer a ta)le o columns #ithin a ta)le, can )ereadily put into #ords, do it"


16/52

When to 6se 'a)les

-ie only signiicant num)ers"


17/52

When to 6se 'a)les

Another ery common useless ta)le is #ord list" 'a)le is an e.ample" 'his inormation could easily )epresented in the te.t"

'hus the rule+ /resent the data in the te.t, or in a ta)le,or in a igure"


18/52

Ho# to Arrange 'a)ular *aterial

Data should )e organiEed so that the like elements readdown, not across"

Compare 'a)le and :" 'hey are eFuialent, e.cept that'a)le reads across, #hereas 'a)le : reads do#n"

'a)le : is the preerred ormat )ecause it allo#s thereader to grasp the inormation more easily"

Words in a column are lined up on the let"


19/52



20/52


'a)le ; is an e.ample o a #ell0constructed ta)le"

It reads do#n, not across"

It has clear heading to make the meaning o the dataunderstanda)le #ithout reerence to the te.t"

It has horiEontal rules 1lines )ut no ertical rules"


21/52

'itles, Kootnotes and A))reiations

'he title o a ta)le should )e concise and not diidedinto t#o or more clauses or sentences"

-ie careul thought to the ootnotes o your ta)le"

I a))reations must )e deines, you can gie all or mosto the deinitions in the irst ta)le" 'hen the later ta)les

can carry the simple ootnote+ “A))reiations as in 'a)le4"%


22/52

-raphs

“A good illustration can help the scientist to )e heard #hen

speaking, to )e read #hen #riting" It can help in thesharing o inormation #ith other scientists" It can help toconince granting agencies to und the research" It can helpin the teaching o students" It can help to inorm the pu)lico the alue o the #ork"%

& *ary Helen Briscoe


23/52

When to Illustrate

Do not use graph i it could )e descri)e in #ords, i"e"only one alue is really signiicant, either a ma.imum orminimum8 the rest is #indo# dressing"

I the choice is graph ersus ta)le, your choice mightrelate to #hether you #ant to impart to readers e.actnumerical alues or simply a picture o the trend orshape o data"

Rarely, there might )e a reason to present the samedata in )oth a ta)le and a graph, the irst presenting thee.act alues and the second sho#ing a trend notother#ise apparent"


24/52

When to Illustrate

Kigure 4 is an e.ample o unneeded )ar graph" 'hisigure could )e replaced )y one sentence in the te.t+

“Among the test group o patients #ho #erehospitaliEed or an aerage o 4@ days, acFuired

inections"%


25/52

When to 6se -raphs

I the data sho# pronounced trends, making aninteresting pictures, use graph"

I the num)er ust sit there, #ith no e.citing trend ineidence, a ta)le should )e satisactory"


26/52

When to 6se -raphs

'a)le 9 and Kigure 2 )oth record e.actly the same data"

(ither ormat is accepta)le, ho#eer Kigure 2 clearlyseems superior #here the synergistic action o the t#o0

drug com)ination is immediately apparent"


27/52

Ho# to /repare -raphs

Kigure 5 is a nice graph" 'he lettering #as large enoughto #ithstand reduction in printing" It is )o.ed, making iteasier to estimate the alues on the right0hand side othe graph"

Do not e.tend the ordinate or the a)scissa )eyond #hatthe graph demand" (.ample, i your data points range)et#een 3 and :;, your topmost inde. num)er should)e ;3 so as no empty spaces in the graph"


28/52

Ho# to /repare -raphs


29/52

7ym)ols and >egends

I there is a space in the graph, use it to present the keyto the sym)ols"

6se only sym)ols that are considered standard and thatare #idely aaila)le"

'he most standard sym)ols are open and closed circles,

triangles and sFuares"


30/52

/hotographs

“>ie is not a)out signiicant details,illuminated in a lash, i.ed oreer"/hotographs are"%

& 7usan 7ontag


31/52

Cropping

7eldom you need the #hole photograph, right out to allour edges"

Crop the photograph to include only the important part"

Crop


32/52

au, '">" ? A)" -hani, A"1234@" 7ediment siEe and depositioncharacteristics in *alaysian ur)anconcrete drains & a case study ouching city" Urban Water ournal ,1114, :@0;9"

A)" -hani, A", Bong, C"H"J" ? >au, '">" 12345"7ediment lushing using tipping lush gate in openstorm concrete drain & a case study in


33/52

Color

I you are considering using color, make sure that thephotographs is still reproduci)le in Black and White"

Color should )e a)le to improe your reportLpaper ratherthan merely decoratie and distracting"


34/52

7ummary & 42 steps to Writing(ectie Results 7ection

4" Determine #hich results to present )y deciding #hich arereleant to the Fuestion1s presented in the Introductionirrespectie o #hether or not the results support thehypothesis1es"

2" rganiEe data in the Results section in either chronologicalorder according to the *ethods or in order o most to leastimportant"

5" Determine #hether the data are )est presented in the orm o

te.t, igures, graphs or ta)les"

@" 7ummariEe your indings and point the reader to the releantdata in the te.t, igures andLor ta)les" 'he te.t shouldcomplement the igures or ta)les, not repeat the sameinormation"


35/52


" Descri)e the results and data o the controls and includeo)serations not presented in a ormal igure or ta)le, iappropriate"

" /roide a clear description o the magnitude o a responseor dierence" I appropriate, use percentage o changerather than e.act data"

:" *ake sure that the data are accurate and consistentthroughout the manuscript"

;" 7ummariEe the statistical analysis and report actual /alues or all primary analyses"


36/52


9" 6se the past tense #hen reerring to your results"

43"


37/52

Discussion

“It is the ault o our rhetoric that #e cannotstrongly state one act #ithout seeming to)elie some other"%

& Ralph Waldo (merson


38/52

Discussion and er)iage

'he Discussion is usually the hardest section to #rite"

(en though the data might )e )oth alid and

interesting, the true meaning o the data may )ecompletely o)scured )y the interpretation presented inthe Discussion"


39/52

Components o the Discussion

'ry to present the principles, relationships, andgeneraliEations sho#n )y the Results" In a gooddiscussion, you discuss & you do not recapitulate & theResults"

/oint out any e.ceptions or any lack o correlation anddeine unsettled points"


40/52

(.ample 0 Discussion

Bong, C"H"J", >au, '">" ? A)" -hani, A" 12345" eriication o eFuations orincipient motion studies or a rigid rectangular channel" Water Science &Technology , 6712, 590@35"


41/52

/oint out e.ceptions

7ho# resultagree orcontradict #ithpreiouslypu)lished #ork"

'ry to present theprinciples sho#n )ythe Results"


42/52

Components o the Discussion

Don$t )e shy8 discuss the theoretical implications o your#ork, as #ell as any possi)le practical applications"

7tate your conclusions as clearly as possi)le"

7ummariEe your eidence or each conclusion"

'he Introduction and the Discussion should unction as a

pair" 'he Introduction should hae posed one or moreFuestions, #hile the Discussion should indicate #hat theindings say a)out the ans#er"


43/52

6se and *isuse o (nglish inDiscussion

“>ong #ords name little things" All )ig things hae little

names, such as lie an death, peace and #ar, or da#n, day,night, loe, home" >earn to use little #ords in a )ig #ay & Itis hard to do" But they say #hat they mean" When youdon$t kno# #hat you mean, use )ig #ords+ 'hey oten oollittle people"%

& 77C B


44/52

*etaphorically 7peaking

Aoid similes and metaphors"

I you use them, use them careully"

Watch out or hackneyed e.pressions, e"g" “timid as amouse%" Interesting and picturesFue #riting results romthe use o resh similes and metaphors8 dull #riting

results rom the use o stale ones"


45/52

*isuse o Words

Watch or sel0canceling or redundant #ords" ("g" “young ueniles%8 “authentic replicas%8 “: a"m" in the morning%8 “ia)le alternatie%"

Amount & 6se this #ord to reer to a mass or aggregate"

6se “num)er% #hen indiidual entities are inoled" “Anamount o cash% is alright" “An amount o coins% is #rong"

Muite &


46/52

'ense

When you state preiously pu)lished indings, you should usepresent tense8 you are reerring to esta)lished kno#ledge"

Nour o#n present #ork must )e reerred to in the past tense"Nour #ork is not presumed to )e esta)lished kno#ledge untilater it has )een pu)lished"

*ost o the A)stract should )e in the past tense, )ecause youare reerring to your o#n results" >ike#ise, the *aterial and*ethods and the Results sections should )e in the past tense,

as you descri)e #hat you did and #hat you ound" n theother hand, Introduction and Discussion should )e in thepresent tense, )ecause these sections oten emphasiEepreiously esta)lished kno#ledge"


47/52

Actie ersus /assie oice

'he actie oice is usually more precise and less #ordythan is the passie oice"

/assie oice unctions #ell in the *ethods section"(lse#here, it rarely should )e used"

(.ample, aoid the passie oice )y saying “'he authors

ound% instead o “it #as ound%"


48/52


49/52

7ummary & 4@ 7teps to Writing(ectie Discussion 7ection

4" rganiEe the Discussion rom the speciic to the general+ yourindings to the literature, to theory, to practice"

2" 6se the same key terms, the same er) tense 1presenttense, and the same point o ie# that you used #hen posing

the Fuestions in the Introduction"

5" Begin )y re0stating the hypothesis you #ere testing andans#ering the Fuestions posed in the Introduction"

@" 7upport the ans#ers #ith the results" (.plain ho# your resultsrelate to e.pectations and to the literature, clearly stating #hythey are accepta)le and ho# they are consistent or it in thepreiously pu)lished kno#ledge on the topic"

" Address all the results relating to the Fuestions"


50/52


" Descri)e the patterns, principles, and relationshipssho#n )y each maor indingLresult and put them inperspectie" Kirst state the ans#er, then the releantresults, then cite the #ork o others" I necessary, pointthe reader to a igure or ta)le to enhance the Ostory$"

:" Deend your ans#ers, i necessary, )y e.plaining )oth#hy your ans#er is satisactory and #hy others are not"

;" Discuss and ealuate conlicting e.planations o theresults"

9" Discuss any une.pected indings" When discussing anune.pected inding, )egin the paragraph #ith the indingand then descri)e it"


51/52


43"Identiy potential limitations and #eaknesses and comment onthe relatie importance o these to your interpretation o theresults and ho# they may aect the alidity o the indings"

44"7ummariEe concisely the principal implications o the indings"

42"/roide recommendations 1no more than t#o or utureresearch" Do not oer suggestions #hich could hae )een easilyaddressed #ithin the study, as this sho#s there has )eeninadeFuate e.amination and interpretation o the data"

45"(.plain ho# the results and conclusions o this study areimportant and ho# they inluence our kno#ledge orunderstanding o the pro)lem )eing e.amined"

4@"In your #riting o the Discussion, discuss eerything, )ut )e

concise, )rie, and speciic"


52/52

All the Best in Nour Report WritingP""

Results Analysis and Discussion Workshop

Documents