making a figure with stata or excel biostatistics 212 lecture 7

Making a figure with Stata or Excel

Biostatistics 212

Lecture 7

Housekeeping

• Lab 5 cleanup– Which p-value is which?– Deciding when to “call” an interaction

• Final Project questions?– Print and hand in to Olivia or Allison (5th floor) by the end of the

day on 9/19/06– 20 points docked for each 1 day late– Email or call for help!

• PLEASE DO COURSE EVALUATIONS– You’ll get a link by email

. cs dead anycac, by(ageover60)

ageover60 | RR [95% Conf. Interval] M-H Weight-----------------+------------------------------------------------- 0 | 3.294296 2.124413 5.108418 11.78094 1 | 3.372508 1.922288 5.916809 9.848343 -----------------+------------------------------------------------- Crude | 4.763402 3.413478 6.64718 M-H combined | 3.329908 2.345418 4.727639-------------------------------------------------------------------Test of homogeneity (M-H) chi2(1) = 0.004 Pr>chi2 = 0.9479

. mhodds dead anycac, by(ageover60)

Maximum likelihood estimate of the odds ratioComparing anycac==1 vs. anycac==0by ageover60

------------------------------------------------------------------------------- ageov~60 | Odds Ratio chi2(1) P>chi2 [95% Conf. Interval]----------+-------------------------------------------------------------------- 0 | 3.343423 31.95 0.0000 2.14383 5.21426 1 | 3.537836 20.94 0.0000 1.98502 6.30536-------------------------------------------------------------------------------

Mantel-Haenszel estimate controlling for ageover60 ---------------------------------------------------------------- Odds Ratio chi2(1) P>chi2 [95% Conf. Interval] ---------------------------------------------------------------- 3.429722 51.90 0.0000 2.400138 4.900967 ----------------------------------------------------------------

Test of homogeneity of ORs (approx): chi2(1) = 0.02 Pr>chi2 = 0.8776

. xi: logistic dead i.anycac*i.ageover60i.anycac _Ianycac_0-1 (naturally coded; _Ianycac_0 omitted)i.ageover60 _Iageover60_0-1 (naturally coded; _Iageover60_0 omitted)i.any~c*i.ag~60 _IanyXage_#_# (coded as above)

Logistic regression Number of obs = 10372 LR chi2(3) = 188.43 Prob > chi2 = 0.0000Log likelihood = -1065.5418 Pseudo R2 = 0.0812

------------------------------------------------------------------------------ dead | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval]-------------+---------------------------------------------------------------- _Ianycac_1 | 3.343423 .7564415 5.33 0.000 2.145898 5.209232_Iageover6~1 | 3.075541 1.040157 3.32 0.001 1.585049 5.96761_IanyXage_~1 | 1.058148 .3922776 0.15 0.879 .5116676 2.188289------------------------------------------------------------------------------

. logistic dead anycac ageover60

Logistic regression Number of obs = 10372 LR chi2(2) = 188.41 Prob > chi2 = 0.0000Log likelihood = -1065.5534 Pseudo R2 = 0.0812

------------------------------------------------------------------------------ dead | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval]-------------+---------------------------------------------------------------- anycac | 3.415374 .6105826 6.87 0.000 2.40583 4.848547 ageover60 | 3.223499 .4453269 8.47 0.000 2.458861 4.225919------------------------------------------------------------------------------

. cs dead anycac, by(male)

male | RR [95% Conf. Interval] M-H Weight-----------------+------------------------------------------------- 0 | 5.901622 3.668887 9.493106 8.3619 1 | 4.287304 2.679298 6.860369 11.74511 -----------------+------------------------------------------------- Crude | 4.763402 3.413478 6.64718 M-H combined | 4.95865 3.549869 6.926513-------------------------------------------------------------------Test of homogeneity (M-H) chi2(1) = 0.883 Pr>chi2 = 0.3473

Housekeeping

• Lab 5 cleanup– Which p-value is which?– Deciding when to “call” an interaction

• Final Project questions?– Print and hand in to Olivia or Allison (5th floor) by the end of the

day on 9/19/06– 20 points docked for each 1 day late– Email or call for help!

• PLEASE DO COURSE EVALUATIONS– You’ll get a link by email

Today...

• Figure basics– Why make a figure?– Types of figures– Elements of a figure

• Steps in making a figure

• Stata versus Excel

• The Final Project, grading

Figures

• Figures are GOOD for presenting overall effects

• Figure are NOT GOOD for presenting specific measurements

Browner, W. Publishing and Presenting Clinical Research

Figures

• “A picture is worth a thousand words”

Figures


52%48%

No Yes

Moderate alcohol consumption in CARDIA participants

How many words is this picture worth?

Figures



48% of CARDIA participants consume alcohol moderately.

Worth = 7 words

Figures



40%

39%

13%

8%

57%26%

9%8%

White Black

0 <1

1-1.9 2+

Alcohol consumption, in drinks/day

Figures



White Black

Drinks/day n=1935 n=1727

0 40% 57%

0.1-0.9 39% 26%

1-1.9 13% 9%

2+ 8% 8%

Worth = 1 small table

Figures



0.0

5.1

.15

.2P

reva

lenc

e of

cor

onar

y ca

lcifi

catio

n

Black women White women Black men White men

By race and genderPrevalence of coronary calcification in moderate drinkers and abstainers

Abstainer Moderate drinker

Figures



% with CAC

Abstainer Mod drinker

Black women .047 .036

White women .054 .049

Black men .068 .132

White men .180 .167

Figures



Worth = “A thousand words”?

Figures



-20

00

-10

00

01

00

02

00

0

Ch

an

ge

in

FE

V1 (

mill

ilite

rs)

0 20 40 60

Pack-years of exposure to tobacco

Menthol smokers Non-menthol smokers

Menthol regression Non-menthol regression

Figures



Worth = 968 data points + lines > 1000 words?

Figures


• Guidelines– Figures should have a minimum of 4 data

points– Convey important effects, or interactions


Figures

• Types of figures– Photographs– Diagrams– Data representation


Figures

• Types of data figures– Pie charts– Bar graphs– Line graphs– Scatter plots– Box plots


Figures

• Elements of a figure– Graphics (non-text)– Labels (axes, lines/bars, etc), other text– Figure legend (Title, explanations, p-values)


Steps in making a Figure

• In Stata:– Sketch the Figure, with title

– Write a do file

– Format so it makes sense and looks nice

– Compose a figure legend

• In Excel:– Sketch the Figure, with title

– Dummy data table in Excel

– Write a do file to fill in table

– Copy and paste the data in

– Format so it makes sense and looks nice

– Compose a figure legend


• Sketch the Figure, with title– Try several versions– Point should be clear at a glance– Requires some artistic vision…


• Can I make this figure with Stata?

Stata vs. Excel for Figures

• Stata– Can create very customizable figures using 1

complex Stata command• Easy to recreate – simple do file• No error• Scatter plots are MUCH easier with Stata

– But…• Harder to create the first time? - no point and click• A little less flexible?


• Excel– Flexible and intuitive point-and-click figures

• Easy to create and modify• Flexible, more options, error bars, adjusted

estimates, etc

– But…• Requires an extra step – copy/pasting to Excel• Harder to reproduce• Much harder to do scatter plots


• Both Stata and Excel can produce very nice-looking figures.


• Write a do file– If making a figure with Stata, your do file

might contain only 1 actual Stata command



might contain only 1 actual Stata commandtwoway (scatter dfev1 cumpy10 if menthol1==1, msymbol(plus) msize(small) mcolor(black)) /// (scatter dfev1 cumpy10 if menthol1==0, msymbol(circle_hollow)) /// (line m cumpy10 if menthol1==1, sort clcolor(black) clpat(dash) clwidth(thick)) /// (line nm cumpy10 if menthol1==0, sort clcolor(black) clpat(solid) clwidth(thick)) /// , ytitle(Change in FEV1 (milliliters), size(large)) yscale(titlegap(5)) /// xtitle(Pack-years of exposure to tobacco, size(large)) /// xscale(titlegap(3)) /// legend(order(1 "Menthol smokers" 2 "Non-menthol smokers" 3 "Menthol regression" /// 4 "Non-menthol regression")) /// scheme(s1mono) /// graphregion(fcolor(none) lcolor(none) ifcolor(none) ilcolor(none)) /// plotregion(fcolor(none) lcolor(none) ifcolor(none) ilcolor(none))



might contain only 1 actual Stata command• Compose using dialog box from menu




– If making it with Excel, you’ll need to produce all the numbers with analysis

• Paste into Excel from log file

• Use Chart Wizard




– If making it with Excel, you’ll need to produce all the numbers with analysis

• Paste into Excel from log file

• Use Chart Wizard

– Either way, you may need additional Stata commands for p-values, figure legend, etc


• Format so it looks nice, and makes sense– With Stata:

• Use dialog box

• Submit, modify, submit again, etc

– With Excel• Point, click, modify


• Compose a figure legend– Title, explanations, p-values, etc– Separate section in manuscript or at bottom of

page – depends on journal


• Example – Excel

• Example - Stata0

.05

.1.1

5.2

Pre

vale

nce

of c

oron

ary

cal

cific

atio

n




Prevalence of coronary calcification in abstainers and moderate drinkers, by race and gender

0

0.05

0.1

0.15

0.2


Pre

vale

nce

of

coro

nar

y ca

lcif

icat

ion


RR of CAC for moderate drinking in Black Men = 1.9 (3.3-8.4)p = 0.017p for interaction = 0.10

twoway (scatter dfev1 cumpy10 if menthol1==1, msymbol(plus) msize(small) mcolor(black)) /// (scatter dfev1 cumpy10 if menthol1==0, msymbol(circle_hollow)) /// (line m cumpy10 if menthol1==1, sort clcolor(black) clpat(dash) clwidth(thick)) /// (line nm cumpy10 if menthol1==0, sort clcolor(black) clpat(solid) clwidth(thick)) /// , ytitle(Change in FEV1 (milliliters), size(large)) yscale(titlegap(5)) /// xtitle(Pack-years of exposure to tobacco, size(large)) /// xscale(titlegap(3)) /// legend(order(1 "Menthol smokers" 2 "Non-menthol smokers" 3 "Menthol regression" /// 4 "Non-menthol regression")) /// scheme(s1mono) /// graphregion(fcolor(none) lcolor(none) ifcolor(none) ilcolor(none)) /// plotregion(fcolor(none) lcolor(none) ifcolor(none) ilcolor(none))

-20

00

-10

00

01

00

02

00

0

Ch

an

ge

in

FE

V1 (

mill

ilite

rs)

0 20 40 60

Pack-years of exposure to tobacco

Menthol smokers Non-menthol smokersMenthol regression Non-menthol regression

graph bar (mean) cac ///

, over(modalc) ///

over(racegender) ///

asyvars ///

ytitle(Prevalence of coronary calcification) ///

title("Prevalence of coronary calcification in moderate drinkers and abstainers", ///

size(medium) span) ///

subtitle("By race and gender", size(medsmall))

0.0

5.1

.15

.2P

reva

lenc

e of

cor

onar

y ca

lcifi

catio

n




Summary figure tips

• Only use a Figure if:– There is an important message to convey– The visual will be more compelling and clear

• Try using both Stata and Excel– Stata will be harder at first, but often worth it– Browner book, Stata book both helpful

• Document, label, and be creative

Final Project, grading

• Grading– 80% (256 out of 319 possible) required to get a

“Satisfactory” score in the class– Also need to turn in all the Labs, even if they

are late


• Final Project will count for 150/319 points– Table – 75 points

• 35 for do file log– Housekeeping commands: open/close log, use dataset, etc

– Analysis: generate numbers in the Table

• 40 for Table itself– Architecture

– Documentation

– Formatting/appearance


• Final Project will count for 150/319 points– Figure – 75 points

• 35 for do file log– Housekeeping commands: open/close log, use dataset, etc

– Analysis

• 40 for Figure itself– Design

– Documentation

– Formatting/appearance


• Extra credit– 10 points extra credit and bragging rights for

the most artistic, creative, and clear figure turned in


• Advice– Find a classmate, give them your Table and

Figure, and get their critiques.• See if they can understand it without any verbal

explanation

That’s it!

• Thanks for your active participation in the course.

making a figure with stata or excel biostatistics 212 lecture 7

Documents

cs dead anycac

mhodds dead anycac

dead odds ratio

odds ratiocomparing

h chi21

odds ratio chi21 pchi2

logistic dead i

h weight