CS599: PPMP + PSP + Project Personal Project Management Process Project Personal Software Process Project using PPMP/PSP PPMP Extra Credit Exercise 7S © 1999 A. Winsor Brown BES/MSEE 599s99PPMP-SS7Sv1.doc–1 v1.0 - 08/09/99 CS599: PPMP + PSP + Project – PPMP Extra Credit Exercise 7S Assignment & Exercise Details Process Exercises Assignment Kit #7 Exercise Evaluation and Grading Process • Specific Items to be Checked per form or Process — Defect Recording Log • PSSP1.1 Exercise Reports – Exit Criteria © 1999 A. Winsor Brown BES/MSEE 599s99PPMP-SS7Sv1.doc–2 v1.0 - 08/09/99 CS599: PPMP + PSP + Project – PPMP Extra Credit Exercise 7S Assignment & Exercise Details Assignment Kit #7S Assignment Kit 7 Contents PSSP1.1 Process Scripts C41 PSSP1.1 Script C42 PSSP1.1 Planning Script C43 PSSP1.1 Development Script C44 PSSP1.1 Postmortem Script C36 PROBE Estimating Script Forms, Templates, and Standards C45 PSSP1.1 Project Plan Summary C47 Task Planning Template C49 Schedule Planning Template C37 Test Report Template C39 Size Estimating Template C27 Process Improvement Proposal C29 Coding Standard C16 Time Recording Log C18 Defect Recording Log C20 Defect Type Standard © 1999 A. Winsor Brown BES/MSEE Instructions Order to submit assignment PSSP1.1 Project Plan Summary n/a Test Report n/a PIP form, including lessons learned n/a Size Estimating Template n/a Task Planning Template n/a Schedule Planning Template Time Recording Log C46 Defect Recording Log C48 C50 C38 C40 C28 n/a C17 C19 n/a Source program listing Other requested materials 599s99PPMP-SS7Sv1.doc–3 v1.0 - 08/09/99 CS599: PPMP + PSP + Project – PPMP Extra Credit Exercise 7S 7S. Relative Error Linear Regression Enhance the linear regression prediction interval spreadsheet 3S, which includes selective masking of outliers, to use relative error instead of absolute error. 1. Spreadsheet 7S Prerequisites and References: DSE Appendix A8 and spreadsheet 3S 2. Spreadsheet 7S Requirements: Write a spreadsheet to calculate an LOC estimate and the 90 percent and 70 percent prediction intervals for this estimate using relative error, not absolute error. .… Also display the relative errror +/90% and +/-70% curves around the regression line; … © 1999 A. Winsor Brown BES/MSEE 599s99PPMP-SS7Sv1.doc–4 v1.0 - 08/09/99 CS599: PPMP + PSP + Project – PPMP Extra Credit Exercise 7S 7S. Relative Error Linear Regression (cont.) 2. Spreadsheet 7S Requirements in detail Write a spreadsheet to calculate an LOC estimate and the 90 percent and 70 percent prediction intervals for this estimate using relative error, not absolute error. The formulae for calculating the linear regression and prediction interval using absolute error are given in Appendix A7 & A8. For relative error, just re-express the dependent variable to produce a new set of dependent variable values, percent error, which you then plug into the formula. Use TREND or LINEST to calculate the regression values. Use the TINV built-in function to calculate the value of the t distribution. You may enhance spreadsheet 3S to develop this spreadsheet. © 1999 A. Winsor Brown BES/MSEE 599s99PPMP-SS7Sv1.doc–5 v1.0 - 08/09/99 CS599: PPMP + PSP + Project – PPMP Extra Credit Exercise 7S 7S. Relative Error Linear Regression (cont.) 2. Spreadsheet 7S Requirements in detail (cont.) Display the relative errror +/-90% and +/-70% curves around the 7S regression line; also include a light vertical line at the "estimated" object LOC and light horizontal lines at the project value and both the sets of UPI and LPI values. © 1999 A. Winsor Brown BES/MSEE 599s99PPMP-SS7Sv1.doc–6 v1.0 - 08/09/99 CS599: PPMP + PSP + Project – PPMP Extra Credit Exercise 7S 7S. Relative Error Linear Regression (cont.) 3. Spreadsheet 7S Testing Thoroughly test the spreadsheet. As one test, use the data for estimated object LOC and actual new and change LOC in table D8 (p 756). Also assume an estimated object LOC value of 386. Under these conditions, the estimated LOC values and the parameter values obtained should be Linear Regression Prediction Interval ß0(-70%) = ~0; ß1(-70%) = ~0.8277; ß0 = 0 ß0(+70%) = 0; ß1(+70%) = ~0.4594 ß1 = 0.5908664 ß0(-90%) = 0; ß1(-90%) = ~1.129; ß0(-90%) = ~0; ß1(-90%) = ~0.4000 t(70%) = 1.108 σ = 187.21923 Projection = 653 t(90%) = 1.860 Range(386,70%) = ~186.946; UPI = ~840.224, LPI = ~466.331 Range(386,90%) = ~311.520; UPI= ~964.799, LPI = ~341.757 This example is worked out in TBD. © 1999 A. Winsor Brown BES/MSEE 599s99PPMP-SS7Sv1.doc–7 v1.0 - 08/09/99 CS599: PPMP + PSP + Project – PPMP Extra Credit Exercise 7S Assignment 7 Details PSSP1.1 Exercise Reports – Exit Criteria 0. The following five items (per paragraphs one and three of C4.5) • Complete process data • Accurate and self-consistent data • Process Report in proper order and format • Historical data used planning • Neat and legible; need not be typed • NO cover sheets, binders, or written reports other than those requested 1. PSSP1.1 Project Plan Summary (per Tables C45 & C46, pg., 690-692). 2. Test Report (per Tables C37..C38, pages 681..682). 3. PIP forms, including lessons learned (per Tables C27&C28, pgs. 668-9) 4. Size Estimate (per Tables C39..C40, pgs. 683..685). 5. Task Plan (per Tables C47&C48, pages 693..694). 6. Schedule (per Tables C49&C50, pages 695..696). 7. Time Recording Log (per Table C17, page 658). 8. Defect Recording Log (per Tables C19 and C20, pages 660-661). 9. Source Program Listing (per personal coding standard - Exercise R3) 10. Other Requested Material • Test Results - Program 7S (per Table D11, pg. 758) • Stu1.XLS with data to date, if at all possible (on diskette, IBM format; all data per instructions through exercise 7S) © 1999 A. Winsor Brown BES/MSEE 599s99PPMP-SS7Sv1.doc–8 v1.0 - 08/09/99