Data Presentation

Lulu Eva RakhmillaEpidemiology & Biostatistics Department

FMUP

Introduction

• When deciding how to display data ask yourself two questions:– What is it that I want to say?– Does my chosen method/ chart/ table

actually show what I want it to show?

Introduction

• Know your audience• Decide what you want to present Are you

presenting data or are you presenting results?

• Tables are good for quantification• Charts are good for illustrating specific points

Why this content necessary

• All the medical sciences, physiology, biochemistry, psychology etc. use quantitative data.

• It is important to be able to manipulate and summarise these data correctly

• For research to be worthwhile, data must be presented meaningfully and correctly interpreted

• This use of figures and tables to make sense of data has a long and distinguished history……

John Snow’s Cholera Incidence map, 1854

• Both the previous charts were drawnwithout the aid of graphics packages

• You have no excuse not to get it right!

• A more recent example of how data display can illuminate an issue…

“Graphic Excellence is that which gives to the viewer the greatest number of ideas in the

shortest time with the least ink in the smallest place.”

Edward R. Tufte

Mediterranean sea

3

4

2

5

1 10

9

7 8

6

Asthma cases in Barcelona by district January 21, 2011

• Distribution by time of day

More informatif...........

Guidelines

• Simplify message without falsifying data• Generally need either graph or table• Present with clarity, brevity• Note prior minor thesis guidelines

Data Presentation Formats

• Tables– Specific data– Exact comparisons between data points

• Bar Charts– Less numerically specific– Examine differences rather than trends/changes– Comparisons of size, magnitude, amounts

• Line Graphs– Not numerically specific– Demonstrate movement, change, trends– Generally over time or concentration

Using a Table

• Should be able to stand on its own• Show data, and possible manipulations

– Percentages, totals, means, averages, ratios, etc.– Columns contain Ind. Variables (that which was

manipulated

Good Table• Legend- complete• Stands on own• Note capitalization• Period after "Table 2"• Units included • Legend above the table; • Note clarifying footnote• Lines of demarcation

separate numerical data from text.

• Gridlines not present

More on Tables• Limit total items/columns

– (But more than than 6-8 datapoints)• No vertical lines• Do not overload with headings• Use captions/footnotes for definitions• Strings of “0’s” or unchanging data might not be

included• Use restraint with decimal places• Obvious abbreviations can be included• Don’t repeat data in text, just call attention to

main points

Preparing a Table• Examine style sheet and examples• One table, one page, double spaced• Use Arabic numerals to number• Group so that comparisons run down column• Logically group data to stress baseline and trends• Round off numbers and align decimals• Create a descriptive caption (no verb required)• Use head- or foot-notes to explain abbreviations• Verify all data• Verify accuracy of use of symbols• Use consistent labeling throughout paper• Proofread carefully

Tables in a Poster/Presentation

• Time limitations- make more simple• Utilize color, shapes, to emphasize• Symbols are okay

Actually Making a Table

• Use publisher’s recommendations• Can use Word or Excel (I like Excel)

Figures

• Illustrations• Photographs• Drawings• Flowcharts

• Line graphs• Bar graphs• Pie charts• Maps

Figures

• Designed to add understanding of information that it difficult to convey with words

• Must be clear, accurate, appropriate• Avoid mere decoration• Need a legend

Parts of a Graph (line)

Line Graphs

• Should have two axes• Y changes as a function of X• Should show data collected at

regular intervals (show trends)• Make curves most bold• Don’t vary line patterns, vary

symbols (color on slides/posters)• Plot the length of intervals so that

slopes are not too steep.

Bar Charts

• One measurable axis• Interval doesn’t matter• Make bars wider than the

spaces between them. • Use color only in slides/posters.

Use conservative patterns for publication

• Show significant differences by letter or asterisk above bars

Scatter Plots

• Examines individual score on two variables.

• Show relationship• Independent Variable

on X (“as a function of”)

Recommendations for Figures

• Read guideline regarding size, color, format!• Is it needed?• Do not have a title• Can it be understood at a glance?• Limit curves (3-5) or bars (6-8; 9-10 grouped)• Plot independent variable on X (time,

concentration), dependent on Y (what happened?)

• Avoid wasted space; legend on field

Recommendations (2)

• Label axes and show units of measure. Use tics and subtics, to not crowd with numbers

• Position, size, shape, length, symbols, angle, color: all are cues. Use when appropriate, and avoid misuse.

• Start scales at “0,” unless you make it clear (tic marks) that you are doing otherwise.

• For a journal, type caption on a separate page so that the figure can be photographed and the type set separately.

Figure Legends• Must accompany Figures.• Should give pertinent, clarifying information

– key to abbreviations– sample size– statistical results– a brief description of how the data were acquired

• Should allow Table/Figure to stand alone• In the legend, both “Table” and “Figure” are spelled

out completely

How to refer to a Table/Figure

• Every table/figure must be referred to in the text• It is best to refer to them in parenthesis:

– Germination rates were significantly higher after 24 hr water soak than in the control (Fig. 1) .

– Note: Fig., here is abbreviated. Not on headings, though.

• Avoid sentences that only direct you to the table: – Table 1 shows the summary results for male and

female height at Bates College.

0

1

2

3

4

5

6

7

10 20 30 40 50 60

2 144677

3 028

4 1

Tabulating and Graphing Numerical Data

Numerical Data

Ordered Array

Stem and LeafDisplay

Histograms Ogive

Tables

41, 24, 32, 26, 27, 27, 30, 24, 38, 21

21, 24, 24, 26, 27, 27, 30, 32, 38, 41

Frequency DistributionsCumulative Distributions

Polygons

O g ive

0

20

40

60

80

100

120

10 20 30 40 50 60

0

5

1 0

1 5

2 0

2 5

3 0

3 5

4 0

4 5

S to c k s B o n d s S a vin g s C D

0

2 0

4 0

6 0

8 0

1 0 0

1 2 0

0 1 0 2 0 3 0 4 0 5 0

S to c k s

B o n d s

S a vin g s

C D

Graphing Categorical Data: Univariate Data

Categorical Data

Tabulating Data

The Summary Table

Graphing Data

Pie Charts

Pareto DiagramBar Charts

Swinscow TDV, Campbell MJ. Statistics at Square One. 10 ed. Sheffield: BMJ Pub; 2002.

Purpose of a research talk

Is not to• Impress the audience• Tell them all you know

about a subject• Present every little

detail of your work

Is to• Give the audience a

sense of what your idea/work is

• Make them want to read your paper

• Get feedback on your work

Know your audience

Know your audience

• Who would be there?– Scientists expert in your field– Scientists not expert in your field– Non experts– Who knows?

Most likely a mix so have something for all

Know your audience

• Keep in mind– They might be tired– They can read – They are thinking “Why should I listen?”– Non-experts will tune off within 2 minutes– Experts after 5 minutes

• What can you do?

What can you do?

• Early motivation - at the beginning of your talk motivate your research with easy to understand examples

• Spoil the punch line - State your results early and in simple terms

• Visuals – Illustrate your idea with images and diagrams

Leave them with these thoughts

• I understood what the problem was and why it was important

• I have an idea of what her solution was and how it was different/better than others

• She knows the literature (i.e. quoted my work ) and we might collaborate on this aspect of her research

Use examples

Examples are your weapon to– Motivate your work– Illustrate the basic intuition– Show your solution in action – Highlight extreme cases or shortcomings

If you are running out of time cut the general case not the example

Where were you?

• People will get lost during your talk, even those who are listening– have a running outline of the main steps of your

idea (more than the talk itself) – use visual clue to highlight where you are in the

process – present it at the beginning of each step

1. Preprocessing2. Filtering3. Texture Extraction4. Decision Trees5. Classification

1. Preprocessing2. Filtering3. Texture Extraction4. Decision Trees5. Classification

1. Preprocessing2. Filtering3. Texture Extraction4. Decision Trees5. Classification

1. Preprocessing2. Filtering3. Texture Extraction4. Decision Trees5. Classification

Related work

• Be familiar with all related work• Don’t list each paper you read• Mainly talk about results that are immediately

related to what you did• References at the end of the talk or better in

the paper itself• Acknowledge co-authors (title slide)

Technical details: in or out?

A fine line– Present specific aspect that show the “meat” of

your work– Leave the rest out. If you were convincing they

will read your paper– Don’t fill up your slides with lots of equations– Prepare back-up slides to answer questions. Leave

them at the end of the presentation

The skeleton

• What is the problem• Motivation and goals• Relevant state of the art• What is your key idea/contribution• Why is your approach good/better• What I just said and what I want to do next

Preparing the presentation

• Less is more. Fill in with narration not words• Use animation sparingly• Use color to emphasize some points but limit to

2 or 3• Be consistent! In the choice and use of color

font size/type etc• Use slide real estate appropriately

Slide layout - Bad

• This page contains too many words for a presentation slide. It is not written in point form, making it difficult both for your audience to read and for you to present each point. Although there are exactly the same number of points on this slide as the previous slide, it looks much more complicated. In short, your audience will spend too much time trying to read this paragraph instead of listening to you.

Slide layout – Good

• Show one point at a time:– Will help audience concentrate on what you are

saying– Will prevent audience from reading ahead– Will help you keep your presentation focused

Fonts - Good

• Use a decent font size• Use different size fonts for main points and

secondary points– this font is 24-point, the main point font is 32-

point, and the title font is 44-point

• Use a standard font like Times New Roman or Arial

Fonts - Bad• If you use a small font, your audience won’t be able to read what you have written

• CAPITALIZE ONLY WHEN NECESSARY. IT IS DIFFICULT TO READ

• Don’t use a complicated font

Color - Good

• Use font color that contrasts sharply with the background

• Blue font on white background

• Use color to reinforce the logic of your structure

• Ex: light blue title and dark blue text

• Use color to emphasize a point• But only use this occasionally

Color - Bad

• Using a font color that does not contrast with the background color is hard to read

• Using color for decoration is distracting and annoying.

• Using a different color for each point is unnecessary

• Same for secondary points

• Trying to be creative can also be bad

Background - Good

• Use backgrounds such as this one that are attractive but simple

• Use backgrounds which are light• Use the same background consistently

throughout your presentation

Background – Bad

• Avoid backgrounds that are distracting or difficult to read from

• Always be consistent with the background that you use

Graphs - Bad

January February March AprilBlue Balls 20.4 27.4 90 20.4Red Balls 30.6 38.6 34.6 31.6

Graphs - Bad

20.4

27.4

90

20.4

30.6

38.6

34.631.6

0

10

20

30

40

50

60

70

80

90

100

January February March April

Blue Balls

Red Balls

Preparing the presentation

• Prepare the slides in advance• Show them to friends• When you think you are done read them again• Check all animations with the sound on

Preparing the presentation

• Practice, practice, practice– Give a practice talk to a general audience– Give a practice talk to an audience of expert– Time your presentation (allow for speed up effect

caused by nervousness)

• Always assume technology will fail you. Have backups.

Delivering the talk

• Be enthusiastic! If you aren’t why should the audience be?

• Make eye contact with the audience• Identify a few “nodders” and speak to them• Watch for questions. Be prepare to digress or

brush off when irrelevant

Delivering the talk

• Point at the screen not the computer• Do not read directly from the PPT or your

notes• Have the “spill” for the first couple of slides

memorized in case you go blank• Finish in time

Handling questions

• Different types – handle accordingly– Need clarification– Suggest something helpful– Want to engage in research dialog– Show that he/she is better than you

• Anticipate questions (additional slides)• Don’t let them highjack the talk (postpone)

How can I get better?

• Practice every chance you can• Observe others

– Steal good presentation ideas– Notice all the things that turned you off

• Seek comments from friends and mentors

Summary

• Data should tell a story• Tailor your presentation to your audience(s)• Use multiple formats to help get your

message to all types of learners• Use images to help explain complex

information

References

1. Swinscow TDV, Campbell MJ. Statistics at Square One. 10 ed. Sheffield: BMJ Pub; 2002.

2. Greenhalgh T. How to read a paper. 3 ed. London: BMJ Pub; 2002.

3. Altman DG, Bland JM. Presentation of numerical data. BMJ. 1996;312:572.