The relationship between landscape preference and looking time: a methodological investigation

This study was an investigation of the relationship between landscape preference and looking time. In the Original Experiment 100 subjects were unknowingly timed while looking at 10 slides of natural landscapes. They then ranked photographic representations of those slides for their preference of them as scenic vistas. A random sample of 25 of the original subjects participated in a Follow-up Experiment about one month later in which they ranked the same landscape simulations as in the first experiment. The analysis of data was divided into two main aspects. In the first aspect the relationship between looking time and preference rank in the Original Experiment was analyzed through summary statistics, frequency distributions, and correlations. In addition, a test to point out the relationship among the average looking times for the slides computed. The second main aspect was an. investigation of the agreement between the rankings in the Original and Follow-up Experiments for the purpose of determining the reliability of the sample subjects’ preferences for the landscapes used as visual stimuli. The major results of this investigation included: (a) the average time spent looking for all the slides was 7.02 seconds, (b) looking-time frequencies varied greatly with the majority of the 100 subjects looking at each slide for approximately 2 to 11 seconds, (c) the overall estimate of correlation between looking time and landscape preference was r = -.073, and (d) of the 25 sample subjects only 2 ranked at least half of the slides (5) in the Follow-up Experiment exactly the same as they did in the Original Experiment; however, the overall estimate of correlation between the two rankings was calculated to be t = .606 (significant at p≤.0001). The investigator concluded that (a) there appears to be no relationship between looking time and preference rank for natural landscapes, (b) there was very little variability among the average looking times for the slides by all the subjects, a range of 6.58 seconds to 7.75 seconds, (c) the subjects’ preference rankings generally ranged across all the integer value ranks for each slide, (d) for all slides except one (I) no integer value for preference rank had a dominate frequency, (e) preference ranking, as was determined by comparing sample subjects’ rankings in the Original and Follow-up Experiments, is a reliable direct method for determining landscape preference, (f) landscape complexity may be a major indicator for determining variation in looking time, and (g) subjects may unconsciously lapse into a rhythmic time-sequence pattern when operating a slide projector.