Designing & Using Charts & GraphsOverviewUse the right toolPie chartsThings to avoid on a piePie ChartSlide 7Proportional bar chartSlide 9Bar chartsSlide 11Line graphsLine graphs continuedSlide 14Picture chart or graphSlide 16Slide 17Tips for diagrams and mapsOrganizational chartsOrganizational ChartFigure and Table LabelsLying with graphsMichigan’s EconomySlide 24Lying charts and statisticsAbsolute valuesAbsolute values vs. percentagesRelativityRelativity IIDesigning & Using Charts & Graphs Compiled by: Jim LucasModified by: Luke ReeseFranklin (42-48, 91-96, 105-114, 280-282, 305-311)OverviewAfter this presentation, you should better understand:Using the right chart or graph to present your informationDesign principles for charts and graphsHow to lie and cheat with graphs, numbers and statisticsUse the right toolPercentage and Percentage and ProportionProportionPie, divided bar chartRankingsRankings Visual table, bar or line graphCumulative totalsCumulative totals Layer or stacked graphValue, time, space Value, time, space relationshipsrelationshipsDiagram, chart, or mapPie chartsPie is based on 100%Slice represents relative amount of a whole Slices should be percentages—not absolute valuesGood design:Easily distinguished pie slices Largest slice starts at 12 o’clock and continues clockwise Slice size decreases clockwise around pieLabel the pie directlyThings to avoid on a pieExploding all the slicesInconsistent labelingLegendsNo labelingPie ChartFigure 1. Michigan’s EconomyPie ChartFigure 1. Industry Contribution to 2007 Michigan’s EconomyProportional bar chartUsed to show proportional amounts as a pie chartUse side scale for more precise measuresMake it wide enough to readIf you have multiple bars, put the least varying factors firstPie ChartFigure 1. Industry Contribution to 2007 Michigan’s EconomyBar chartsUse to show change in ordinal (ordered sequence, e.g., 1st, 2nd, 3rd) or nominal (name or form, e.g., Yahoo, Google, Ask.com) dataUsually vertical, but use horizontal if it makes senseColor code consistentlyAvoid over emphasizing one lineIf you use a legend, have it follow the order of your barsFigure 2. 2006 cattle loss percentages by cause Figure 3. 2006 cattle loss percentages by stateLine graphsUsed to show change in trends and continuous dataVary color or line weight to call attention to specific pointsPoints on the line are important for precise valuesBe consistent with labelingLine graphs continuedPut warm, bright colored lines on topUse a scale that accurately reflects the change in informationFigure 4. Currency Exchange Rate (US$ to NZ$ and AU$) for first day of month, 2006Picture chart or graphYou can use pictures in the place of a bar chart or graphFollow color and theme conventionsHave the pictures align in the same directionVary size not color or shadingPalm Tree GrowthPalm Tree GrowthTips for diagrams and mapsFollow the culture of the informationIdentify relationships on a chartUse color, line weight, shading to organize informationShow only relevant informationAvoid gridlines unless neededOrganizational chartsSquares or rectangles to indicate divisionsTop down structureSolid lines - direct relationshipsDashed line - indirect relationshipsOrganizational ChartAdministrativeAssistantStaffFacultyAssistantChairSpecialistsTechnologyManagerInformation OfficersCommunicationsManagerDepartmentChairpersonFigure 5. Departmental organizational chart, February 2008Figure and Table LabelsVery importantNumber each figure sequentiallyPlace label under chart in same location (repetition)Provide a titleProvide an informative caption to identify the purpose or meaningAll figures must be referenced in the narrativeFigure placed as close as possible immediately after the narrative referenceLying with graphsLying with graphsAltering the y-axisDistorted shapesUneven start point Using color, lines or other technique to distort differences Using 3-D effects or distortionsMichigan’s EconomyCattle Loss in MichiganDiseasePredationOther020406080100120% LossLying charts and statisticsLying charts and statisticsPercentages and absolute valuesNumbers are relativeSquishy numbers and figuresValue labelsErrorAbsolute valuesNumber of cattle deaths attributed to coyotesNMontana 2,000Michigan 40Absolute values vs. percentagesCattle deaths attributed to coyotesN % of total deathsMontana 2,000 10%Michigan 40 20%RelativityCattle deaths attributed to coyotes over 5 yearsN % of total deathsMontana 2,000 10%Michigan 40 20%Relativity IICattle deaths Cattle deaths attributed to attributed to coyotes over 5 coyotes over 5 yearsyearsN Monetary Loss % loss of total salesMontana 2,000 $500 9%Michigan 40 $300