A Critical Review of Line Graphs in Behavior Analytic Journals

1 Review ARTICLEA Critical Review of line Graphs in BehaviorAnalytic JournalsRichard M. Kubina E. Kostewicz2&Kaitlyn M. Brennan2&Seth A. King3 Published online: 3 September 2015#Springer Science+Business Media New York 2015 AbstractVisual displays such as Graphs have played an instrumental role in psychology. Onediscipline relies almost exclusively on Graphs in both applied and basic settings, behavioranalysis. The most common graphic used in Behavior analysis falls under the category of timeseries. The line graph represents the most frequently used display for visual analysis andsubsequent interpretation and communication of experimental findings.

2 Behavior analysis, likethe rest of psychology, has opted to use non-standard line Graphs . Therefore, the degree towhich graphical quality occurs remains unknown. The current article surveys theessentialstructureandquality featuresof line Graphs in behavioral Journals . Four thousand threehundred and thirteen Graphs from 11 Journals served as the sample. Results of the surveyindicate a high degree of deviation from standards of graph construction and proper labeling. Adiscussion of the problems associated with graphing errors, future directions for graphing inthe field of Behavior analysis, and the need for standards adopted for line Graphs construction standardsBehavior analysis, a subfield of psychology, owes a great debt to the visual display of data.

3 Forexample, the cumulative recorder offered a standard visual display of an organism sperfor-mance data. The distinctive visual patterns of Behavior led to the discoveries such as schedulesof reinforcement (Lattal2004 ). As Behavior analysis moved forward in time, the visual displaysshifted from cumulative recorders to line Graphs . Data show that cumulative records in theEduc Psychol Rev (2017) 29:583 598 DOI *Richard M. Kubina, Education Program, The Pennsylvania State University, 209 CEDAR, Building UniversityPark, State College, PA 16802-3109, USA2 University of Pittsburgh, Pittsburgh, PA, USA3 Tennessee Technological University, Cookeville, TN, USAJ ournal of the Experimental Analysis of Behaviorcontinue to appear infrequently and in otheryears not at all (Kangas and Cassidy2010 ).

4 The shift away from cumulative records to line Graphs coincided with the advent of anemphasis on applied work. The oft-cited paper of Baer et al. (1968 ) laid the foundation fordiscerning the facets of Behavior analysis. Three of seven characteristics of applied behavioranalysis have a direct link to visual displays of data. First,Analyticrefers to a convincingdemonstration of an experimental effect. The preferred medium for all analysis of data occursthrough Graphs . Second,Effectiveconveys the requirement for the intervention to produce apractical and meaningful magnitude of Behavior change.

5 line Graphs allow for the determi-nation as well as the public documentation and communication of the significance of behav-ioral improvements (Spriggs and Gast2010 ). And third,Generalitymeans that the behaviorpersists across time, environments, and operant responses within a class. The line graph , partof the family of time series Graphs , directly portrays the extent to which Behavior does or doesnot of graphic presentation for line Graphs have quality standards necessaryfor the accurate representation of data. A number of publications have described thestandards for proper construction for line Graphs (American National Standards Insti-tute and American Society of Mechanical Engineers1960, 1979; American StandardsAssociation1938; American Statistical Association1915; Department of the Army2010).

6 For example, the publication ofTime series charts: a manual of design andconstructionset forth agreed upon standards for line Graphs (American StandardsAssociation1938). The committee provided guidance on many specific featuresranging from scale rulings and graph dimensions to the weight of lines and use ofreference symbols. Through time, many professional organizations and researchershave continued to offer principles of design and procedure for constructing high-caliber line Graphs ( , Behavior Analysis Cooper et ; Statistics Cleveland1993,1994; General Science Scientific Illustration Committee1988; TechnicalDrawing, Drafting, and Mechanical Engineering Giesecke et ).

7 Table1 listsmajorquality featuresof line Graphs tailored toward use in the behavioral analysis of the following basic Behavior analysis (Alberto and Troutman2013 ;Catania1998 ; ; Malott and Shane2014 ; Mayer et ; Pierce and Cheney2013 ;Vargas2013 ) and single case design books (Barlow et ;Gast2010 ; Johnston andPennypacker2009 ;Kazdin2011 ;Kennedy2005 ) corresponds to the graphical standards for aline graph previously listed. In addition toquality standardsline Graphs have anessentialstructureconsisting of two axes, the horizontal and vertical, representing a time unit and aquantitative value, respectively.

8 Time units can cover minutes, hours, days, weeks, and yearsbased on the second (National Institute of Standards and Technology2014 ).The range ofbehavior on the vertical axis spans dimensionless quantities like percentages and ratios todimensional quantities such as repeatability and temporal extent measured with frequency andduration, respectively (Johnston and Pennypacker2009 ).Not adhering to theessential structuremay yield distorted, exaggerated, or impreciseinformation. Theessential structureshows change over time. Figure1 shows three line graphswith the same data.

9 The first line graph made following theBproportional construction ratio,^discussed later, displays a series of data with a moderately increasing variable trend. The linegraph has an extended vertical axis changing the variability from moderate to low. The trendalso increases when compared to the previous graph . Stretching the horizontal axis in the thirdline graph depresses the trend and decreases Psychol Rev (2017) 29:583 598 Table 1 Quality features of a line graph and measurementEssential structureFunctionMeasuredVertical axis labeled with quantitative measure;horizontal axis labeled with time unitTo show the change in the measure over time(Harris1999 )What label does the vertical and horizontal axes maintain?

10 Does the figure maintain a line for each axis?Quality featureVertical axis length has a 2:3 ratio to thehorizontal axisTo properly display data variability and limit distortion(Cooper et ; Parsonson and Baer1978)Is the ratio of vertical to horizontal axis 5:8 to 3:4 (63 to 75 %difference)? Do axes line up in all multiple baseline Graphs ?Are axis lengths the same for all figures with the same unitwithin each article? Are all figures with the same unit scaledto the same minimum and maximum?Tick marks point outwardTo prevent or minimize data obfuscation (Cleveland1994) Onbothaxes,aretickmarkspointing outward for the entire lengthof the axes?