Visual Analysis of Graphic Data

Description of the Strategy

The visual analysis of graphically displayed data is the primary strategy applied researchers use for analyzing data collected within the context of a singlesubject research design. Unlike group research methodology, in which participants' data are grouped and statistically analyzed, single-subject research methodology emphasizes the importance of analyzing each participant's performance across experimental conditions to determine the effect the independent variable has on each individual's behavior. Simple line graphs are used to visually display primary data, and visual analysis guidelines are followed to increase accuracy in interpreting data patterns.

The graphic display and visual analysis of each research participant's performance data have several advantages. Specifically, the ongoing and formative evaluation of participant behavior is helpful in determining whether to maintain, modify, or replace an intervention under investigation. Research-based data-decision rules can be applied by using line graphs and adherence to visual analysis guidelines. Single-subject research reports neither underestimate nor overestimate the effectiveness of an intervention, since each research participant's data are graphed and analyzed. Serendipitous findings are also possible that may lead to new and interesting lines of research. The graphic display of primary data for each research participant permits the independent analysis of data patterns both within (intrasubject replication) and across (intersubject replication) participants, allowing for an independent determination as to the internal and external validity of the findings. Visual analysis guidelines may be used to analyze group data, provided the target behavior is measured repeatedly and the data are plotted on a line graph. In such cases, separate graphs should be presented for individual participants, in addition to the group data.

To understand the visual analysis process, it is important to understand the ABC notation system commonly used with single-subject research designs. The use of this system permits single-subject design researchers to convey information regarding their analyses in a concise manner. The capital letter A is used to refer to the baseline or probe condition during which the target behavior (dependent variable) is repeatedly measured under preintervention conditions. The baseline, or A condition, may represent the “natural” frequency, duration, magnitude, and so forth of the target behavior. The A condition is simply the first condition of the investigation to which subsequent participant performance is compared. The capital letter B is used to refer to the first experimental condition introduced in the investigation, and it represents the instructional procedure or intervention (independent variable) under study. If additional experimental conditions are introduced into the investigation, each will be assigned a letter according to order of introduction (e.g., C, D).

When referring to some single-subject research designs, such as withdrawal, reversal, and multitreatment designs, a hyphen is inserted between letters separating adjacent conditions. For example, a [Page 1596]withdrawal design would be depicted as an A-B-A-B design, and a multitreatment design in which a baseline condition is followed by the alternation of two experimental conditions would be depicted as an A-B-C-B-C design. When two experimental conditions (B and C) are combined to build a “treatment package,” the hyphen between letters is omitted (BC), thereby reflecting the combining of the two experimental conditions into one condition (e.g., B = contingent token reinforcement, C = response cost). A possible single-subject research design for evaluating the effects of a token reinforcement procedure (B) alone, compared to a token reinforcement procedure plus response cost procedure (BC), would be depicted as an A-B-BC-B-BC design. Primes (B′, B″) are sometimes used to show slight changes in the same experimental condition, such as when thinning the schedule of reinforcement from continuous reinforcement (B) to fixed ratio-2 (B′) to variable ratio-3 (B″). When conducting a visual analysis, it is important to indicate whether the first, second, or third introduction of the A condition is being referenced or the first, second, or third introduction of the B condition. This indication is done by placing a numeral at the lower right of the letter (subscript) indicating whether it is the first, second, or third introduction of the condition (A1-B-A2-C1-A3-C2).

...