The Use of Importance Performance Analysis in the Hospitality Industry

Changes in Tourism Industry Quality of Work Life Practices

“…Martilla and James (1977) suggested mean importance values should be plotted against mean performance values, and others have subsequently followed this approach (Crompton & Duray, 1985). Grand mean (mean of all mean scores) was the most frequently used technique (34%) of 42 IPA studies analyzed (Janes & Wisnom, 2003). Janes and Wisnom (2003) further found, however, that grid line placement was often determined in various ways.…”

“…Grand mean (mean of all mean scores) was the most frequently used technique (34%) of 42 IPA studies analyzed (Janes & Wisnom, 2003). Janes and Wisnom (2003) further found, however, that grid line placement was often determined in various ways. Twenty-eight percent plotted the grid lines based on researcher judgment, 25% used midpoint of the response range, and 12% used median plots.…”

“…The IPA involves three steps: 1) the selection of the attributes to be measured, 2) the measurement of importance and performance for each attribute, and 3) the plotting of the average score given to each attribute on a two-dimensional action grid. According to Janes and Wisnom (2003), the IPA has been used throughout the last three decades by hospitality related researchers as a tool to evaluate the views of staff, facilities, areas or programs. Most often self-administered survey instruments have been distributed after an experience transpired.…”

“…On average, 30 attributes were evaluated and placed on an action grid (matrix) to present the findings. The overall mean was used most often to determine which of the four categories an attribute was placed (Janes & Wisnom, 2003).…”

“…Martilla and James (1977) indicate that separating the importance and performance measures helps to minimize compounding and order effects. Janes and Wisnom's (2003) analysis of 42 studies using the IPA technique suggest that most all did this by placing the questions at two different points in the interview or questionnaire. Only three of these studies measured the attributes at two completely different times (before and after the experience) as Martilla and James (1977) suggest.…”