Describe the various methods for estimating task duration and discuss how these methods improve the accuracy of task duration estimates, taking into account the variation that is present in task duration. Provide practical examples to help illustrate the differences between the methods. Wysocki (2009, P149) indicates that there are “Six Methods for Estimating Task Duration”, these are 1. Similarity to other activities 2. Historical Data 3. Exert Advice 4. Delphi Technique 5. Three Point Technique 6. Wide-band Delphi Technique(Wysocki 2009, p. 149)
Before the overall project duration can be estimated in must first be broken down into manageable smaller activities, the series of tasks generated, once identified and organized into a logical sequence this forms the basis of the Project Work Breakdown Structure (WBS). Once the WBS is built the activities can be assigned durations based on the various estimating methods to develop the overall project plan. 1. Similarity to other activities. If the task is something simple and/or a repetitive task that has been done in previous projects on numerous occasions it is safe to say that the durations taken previously could be used again.
Whilst this is probably the most simple approach other considerations must also be looked at, for example there is no point in estimating that a taxi duration from the city centre in Jakarta to the airport will take 30minutes because that is what it took the last time. The environment, time of day, weather conditions and many other considerations must be factored in to ensure the time in realistic. 2. Studying Historical Data. If in a project we had to plan for many taxi trips to the airport as mentioned above perhaps considering the historical data of durations would give us a better average time.
So in the last year I have made 12 trips to the airport (1 a month) and at various times of the day and night, the durations have varied from 30mins to 3hrs thus for our project plan we use the average of all these trips (e. g. 2 hrs). 3. Seeking Expert Advice. Again using the taxi example, when phoning and booking the taxi it is always worthwhile asking the company what their advice would be in terms of pick up time if we have to catch a flight at 4pm. In this case they will only provide an estimate based on the day and time we have given them however they too may carry a database of average times that could be sought / requested. 4.
Delphi Technique. By using the Delphi Technique a collective approach is used where several team members may be asked to estimate the duration of the taxi journey, once the estimates are made the outer quartiles are challenged and then a re-estimate is made, a third round of challenging and estimating is then made with the average of this estimate used in the plan. In my simple taxi example this could be a good or bad approach depending on the team members. Asking 6 local Jakartan Taxi Drivers may result in a robust estimate whilst asking 6 London Cabbies may be significantly out thus the need for experienced people will help in this process. . The 3 Point Technique. The 3 point technique is a probabilistic estimate taking account of optimistic, pessimistic and likely durations and using a simple mathematical formula to calculate an estimated duration. The downside to this technique is if the durations are widespread it is difficult to identify the “likely” duration hence implementing the formula is difficult. For the taxi example this is probably an over complex solution to a simple task. 6. Wide Band Delphi Technique.
A variant of the Delphi and 3 point technique where the collective team members in the Delphi Technique are asked to provide optimistic, pessimistic and likely durations per the 3 point technique The final 3 methods are more complex and time consuming thus would likely not be employed in the simple example above, however for longer duration, complex projects with large WBS’s there may be a benefit in these estimating techniques. References: Robert K. Wysocki, Effective Project Management, John Wiley & Sons (P&T) p. 149