12 Making process predictions

12.3 Models of effort and cost, cont.2

The Putnam´s SLIM model was constructed for US Army use in 1978 to cover projects exceeding 70 KLOC. Putnam´s model assumes that the effort for software development projects is distributed similarly to a collection of Rayleigh curves, one for each major development activity, see Figure 12.6.

Like the COCOMO model it is based on empirical studies. It relates size (S), a technology factor (C), total project effort in person years (K), and elapsed time to delivery in years(t_d) (in theory the value for which max of the Rayleigh curve is obtained):

S = C K^1/3 t_d^4/3 The equation allows you to assess the effect of varying delivery date on the total effort needed to complete the project. So for instance a 10% decrease in elapsed time S = C K^1/3 t_d^4/3 =C K´^1/3(0.9 t_d)^4/3 results in K´/K = 1.52, i.e., in a 52% increase in total life-cycle effort.

12.4 Problems with existing modeling methods

Surveys reveal that rarely PRED(0.25) exceeds 0.75, i.e., the studies paint a clear picture of model insufficiency. Many studies agree that b in the effort-duration models is about 1/3. However, there is little concensus about the effect of reducing or extending duration. Models with many parameters (eg. cost drivers) are not necessarily preferable (accuracy, subjectivity, independency, static). Size estimates in LOC are not available early in the process.

12.5 Dealing with problems of current estimation methods

• use local data definitions
• calibrate models in the actual environment
• use independent estimation group
• reduce input subjectivity
• do preliminary estimate (group, like Delpi; by analogy) and re-estimation
• use other than LOC for early size estimation
• use locally developed cost models
• use PRED(0.25) or other to assess acceptable accuracy
• see Table 12.7