[Priming effects in picture problems: preliminary solutions]

Two experiments explored whether picture puzzles are an adequate instrument to investigate implicit memory for pictorial information. During the testing phase, the subjects had to identify hidden figures in picture puzzles. In a preceding learning phase, the priming conditions were varied systematically. In the first experiment, some subjects had to solve picture puzzles in the learning phase, whereas others made esthetic judgments (global processing) or estimated the number of triangles in the picture puzzles (local processing). In the second experiment, the subjects inspected copies of figures that were hidden at testing, modified versions of these figures, or their names in the learning phase. In the first experiment, the subjects of the different encoding conditions showed comparable priming effects. Picture puzzles that had already been processed or seen during learning were solved more often than new ones. Interview data revealed that subjects in the local or global processing conditions did not identify hidden figures at encoding. Furthermore, these subjects could not discriminate between old and new picture puzzles in a final explicit test of recognition. Thus, nonconscious storage of perceptual information that is not semantically interpreted may be sufficient to evoke priming effects. In the second experiment, the subjects in the different encoding conditions showed reliable priming effects, too. The presentation of the duplicates at encoding produced the greatest amount of priming. Effects of verbal priming, however, indicate that the solution of picture puzzles is not based solely on perceptual information. Depending on the priming conditions at learning, the solution of picture puzzles may be based primarily on data-driven processing or may be guided more heavily by conceptual information. It is concluded that perceptual, lexical, and/or conceptual information can contribute to the solution of picture puzzles.