Introduction: Many tactile graphics designed for individuals with visual impairments consider single factors. According to the results of our previous study, there may be interactions among scale, representation, and complexity factors. We conducted this integrative study with these three factors. Additionally, for the representation factor, we introduced a new level that mixed the two common levels (line drawing, (LD), and texture picture, (TP)), into a textured-line drawing, (TLD). Methods: We included 18 participants with congenital blindness. They were asked to identify and name tactile graphics. The design of tactile graphics involved three factors, each at different levels, including scale (large, medium, and small), representation (TP, LD, and TLD), and complexity (easy and complex). We recorded the identification time and accuracy and conducted three-way analysis of variance to investigate interactions. Results: The identification time for small-scale graphics was shorter than that for large-scale graphics. The accuracy for small-scale graphics was higher than medium-scale graphics. Under the TLD mode, the accuracy for small and medium-scale graphics was higher than that for large-scale graphics. For medium-scale graphics, TLD performed better than LD. Discussion: Because the sizes of small-scale graphics were similar to those of actual objects, they were easy to identify. If the TLD mode is used for medium-scale graphics, the components in the operation area can be completely presented, which is helpful for identification. However, if large-scale graphics are used under the TLD mode, the operation area is relatively small and difficult to identify. Implications for Practitioners: It is recommended to present objects at 1:1 under the small-scale. Under the medium-scale, the operation area can be presented by closed plains, while non-operation areas can be presented using lines. Under the large-scale, it is recommended to reduce the size of the graphics to an extent where both hands can be used to explore.