Page 260 - Industrial Process Plant Construction Estimating and Man Hour Analysis
P. 260

Statistical applications to construction Chapter  12 239



                Histogram
             Use Excel’s chart capabilities to plot the graph for the histogram.
                To use the Excel chart capabilities, highlight range D77-E79, and select
             Insert, Charts, Insert Scatter Chart, Insert Combo Chart, Create Custom
             Combo Chart, Cluster Column, and Chart Elements: Axes, Axes Titles,
             Chart Title, Gridlines, and Data Table
                         Bin                            Frequency
                         2.00                           1
                         4.00                           2
                         More                           1

                                        Histogram
                     3
                   Frequency  2
                     1
                     0
                             2.00         4.00         More         Series1
                  Series1     1            2            1
                                        Bin

                Graphical analysis of process piping and equipment…Validate direct
                craft man-hour tables
             The practical examples that follow for process piping and equipment use graphic
             analysis to plot the data. Plotting the data is useful to see and understand the rela-
             tionship between two-variable data and allows results to be displayed in pictorial
             form. Thisprovides insight into the data set to help with testingassumptions, model
             selection, regression model validation, relationship identification, factor effect
             determination, and outlier detection. Plotting of data allows the estimator to check
             assumptions in statistical models and communicate the results of an analysis.

             12.3 Graphical analysis of process piping


                Graphs
             A graph is a pictorial illustration of the relationship between variables.
                Assume for historical man-hour data that x is the independent variable and y
             is the dependent variable:

                Straight-line graph
                                            ð
                                                    ð
                            y ¼ a + bx; Y ¼ a + y y1Þ= x x1Þ xðÞ
             where y is the dependent variable;
                a is the intercept value along the y axis at x ¼ 0;
                b is the slope, or the length of the rise divided by the length of the run,
                b ¼ (y   y1)/(x   x1);
                and x is the independent or control variable.
   255   256   257   258   259   260   261   262   263   264   265