Efficiency of information transmission at different types of graphical representations of quantitative data

Đuričić, Mladen. (2012). Efficiency of information transmission at different types of graphical representations of quantitative data. Diploma Thesis. Filozofski fakultet u Zagrebu, Department of Psychology. [mentor Ivanec, Dragutin].

PDF (Croatian) - Registered users only Download (1MB) | Request a copy

Abstract

Conducted study was about the accuracy of size estimations for three different types of graphs (pie, column and bar) with different number of shown values (2, 3 and 5) and dimensions of the display (2D/3D) in the factorial experimental design (3 x 3 x 2). The procedure consisted of the projection of slides with graphs to groups of participants who estimated the size of the presented values by dividing the line in the answers protocol to as many segments as values presented on the graph, while the size ratio on the line should be as similar as possible to the size of the displayed values. The participants gave their answers during graph were presented and short pause before a new task. The result was a total estimation error for all values shown on the each graph. The results were processed with repeated measures ANOVA. Results show the existence of a statistically significant interaction between all three variables (the chart type, size and number of dimensions), and between number of values with type of graph, and between number of values and dimensions. Our study also found that participants gave significantly more accurate estimates for graphs with presented 2 and 3 values compared to those with 5 values. Other two variables (graph type and dimensions), and their interactions were not provide statistically significant results.

Actions (login required)

View Item