Selecting Improvement Subjects Using Pareto Charts and Stratification Starlight Homes Inc. is a building contractor specializing in upscale homes in the Southwest. Before each new home is sold, Starlight conducts a final inspection of the home and repairs any defects. In addition, Starlight receives a "punch list" of defects (to be corrected) compiled by the buyers following the sale. Ricardo Alvarez, Starlight's lead supervisor, has concluded that even if it should cost more to do the work right in the first place, it will be a lot cheaper than going back later to fix the defects. To reduce costs, improve the quality of Starlight's homes, and reduce the number of complaints after the sale, Ricardo has assembled data from his final inspections and the punch lists for the last 20 homes sold. These data are listed in the following chart:

Defect Type Occurrences Defect Type Occurrences Damaged walls 13 Doors 14 Exterior paint 5 HVAC 11 Plumbing 33 Roof 3 Caulking 28 Masonry 2 Electrical 25 Interior paint 61 Cabinetry 12 Landscaping 16 Woodwork 46 Fixtures 7 Construct a Pareto chart to illustrate the defect types by several defects. Which two defect types appear to be the most significant? Should Ricardo focus his attention on these two categories of defects? Ricardo decided to stratify the Pareto chart by cost before making any decision. His bookkeeper developed the average cost per repair event per category. The data are shown here:

Average Cost per Repair Event Defect Type Average Cost Defect Type Average Cost Damaged walls $126 Doors $11 Exterior paint 25 HVAC 110 Plumbing 78 Roof 72 Caulking 7 Masonry 290 Electrical 74 Interior paint 4 Cabinetry 88 Landscaping 34 Woodwork 5 Fixtures 31 The second-level chart developed from these numbers should show Ricardo which two or three defect types are the most significant in terms of cost, enabling him to put his efforts where they will do the most good. What are they? Was Ricardo justified in going to the second level (of stratification) before deciding where to focus his efforts? Why?

Constructing a Cause-and-Effect Diagram Your team has been given the charter to make recommendations for improving the cleanliness of the company's restrooms. The team has finished compiling a list that it considers being the possible contributors to less than desirable restroom cleanliness:

Janitor service Use of paper towels Lack of paper towels Slobs User carelessness User attitude Janitor attitude Paper dispensers Poor lighting Too small Unreliable plumbing Type of floor material Janitor pay Management inattention Janitor supervision Air conditioning Insufficient exhaust In-house plumbers Cleaning materials Cleaning equipment Paper receptacles Maintenance Cleaning schedule Leaking faucets Janitor procedures Construct a cause-and-effect diagram incorporating all these ideas.

Translating Your Histogram Your process has a normal histogram with located at 10 millimeters, 1 points at 9.9 and 10.1 millimeters, 2 points at 9.8 and 10.2 millimeters, and 3 points at 9.7 and 10.3 millimeters.

If your customer will accept parts measuring between 9.7 and 10.3 millimeters, how many parts of every 1,000 produced would you expect to scrap? If your customer notified you that henceforth it would accept only parts between 9.9 and 10.1 millimeters, what would you anticipate your scrap rate to be? What would have to be done to bring the scrap rate at this new customer requirement back down to what it had been at the former customer specification? Compared with the bell curve above, what would the new histogram have to look like?