Home > Articles > Software Development & Management

  • Print
  • + Share This
This chapter is from the book

Step 3: Defining Data Collection

After choosing the initial set of measures, the data necessary to support the measures can be defined. During this step, the focus should be on the following:

  • Data definitions
  • Data collection points
  • Data collection responsibilities
  • Data collection vehicles

Each of these will be discussed in detail.

Data Definitions

Each piece of data necessary for a measure needs to be identified and defined in terms everyone can understand. For example, if the productivity measure of function points per hour is selected, function points and effort will be required. Effort needs to be defined based on what activities to include (requirement defin-ition, design, coding, and so on). If an organization is planning to benchmark against industry data, identifying what activities are included in the industry numbers will ensure that accurate comparisons can be made. At a minimum, telling the benchmarker what data has been collected will be necessary so that the benchmark database can be filtered appropriately. The definitions may at first seem obvious, but it is amazing how different definitions can be. I have seen major discussions break out about defining a project completion date.

Data Collection Points

Data collection activities must be integrated with the development life cycle so the measurement becomes part of the process and is not perceived as something extra. Data should be collected only at the points necessary to support the selected measures. For example, if the goal is to improve project productivity, and the measure is function points per hour, you will need to count function points at implementation time. Effort, on the other hand, should be collected throughout the project life cycle. And if the goal is to manage change of scope in function points, the function point counting activity will be required multiple times during the project life cycle.

Data Collection Responsibilities

To ensure collection of the data, you need to define roles for collecting and reporting each piece of data. Various individuals may be responsible for recording, collecting, and reporting the data. Knowing the data definitions and collection points will aid in determining the most appropriate assignment of personnel for data collection and reporting responsibilities. The approach chosen for establishing measurement of staff resources may also have an effect. These approaches are discussed later in the chapter.

Data Collection Vehicles

Wherever possible, utilize existing data collection forms or systems. Avoid reinventing the wheel; this just adds additional overhead and confusion to the process. There are automated tools that can help with some of the data collection. However, knowing the data collection requirements is important prior to selecting a tool. While defining the forms or methods of data collection, you also need to define the process of providing the data to the measurement personnel.

Examples of data definition, data collection points, data collection responsibilities, and data collection vehicles are documented in Table 6-2 in the Data Definition and Responsibility columns.

Now that the initial measures and required data have been defined, the means of reporting the data can be developed.

Table 6-2

Goal/Initiative

Measures

Data Definition

Responsibility

Improve project productivity (Goal)

Function points per hour

Function points counted at project implementation

Effort captured throughout proj ect life cycle

Function point specialists along with subject matter experts (record data in a spreadsheet)

Developer (records data in time entry system)

Improve project quality (Goal)

Delivered defects per function point

Function points counted at project implementation Delivered defects after implementation for three months

Function point specialists along with subject matter experts (record data in a spreadsheet)

Help desk (records data in defect tracking system when reported by users)

Reduce project cost (Goal)

Cost per function point

Function points counted at project implementation Labor cost calcu lated based on effort Non-labor cost captured through out the project

Function point specialists along with subject matter experts (record data in a spreadsheet)

Project manager (calculates and records data on project completion form)

Implement formal Inspections (Initiative)

Defect removal efficiency rate(defects found prior to implementation divided by total defect found)

Delivered defects per function point

Function points counted at project implementation

Defects found prior to implementation

Delivered defects after implementation for three months

Function point specialists along with subject matter experts(record data in a spreadsheet)

Inspectors and measurements analysts (record data on defect form and in a spreadsheet)

Help desk (records data in defect tracking system when reported by users)


  • + Share This
  • 🔖 Save To Your Account