1. Halstead Metrics

The metrics derived from Halstead measures can be used to estimate testing efforts, as given by Pressman. Halstead developed expressions for program volume V and program gram level PL, which can be used to estimate testing efforts. The program volume describes the number of volume of information in bits required to specify a program. The program level is a measure of software complexity. Using these definitions, Halstead effort e can be computed as:

$\begin{aligned} P L &=1 /[(n 1 / 2) \times(N 2 / n 2)] \\ e &=V / P L \end{aligned}$

The percentage of overall testing effort to be allocated to a module k can be estimated using the following relationship:

Percentage of testing effort $(k)=e(k) / \Sigma e(i)$

where $e(k)$ is the effort required for module k and $\Sigma e(i)$ is the sum of Halstead effort across all module of the system.

2. Development Ratio Method

This model is based on the estimation of development efforts. The number of testing personnel required is estimated on the basis of the developer-tester ratio. The results of applying this method is dependent on numerous factors including the type and complexity of the software being developed testing level, scope of testing, test-effectiveness during terror tolerance level for testing, and available budget.

Another method of estimating tester-to-developer ratios, based on heuristics, is proposed by K Iberle and S. Bartlett.This method selects a baseline project(s), gathers testers-to-developer ratios, and collects data on various effects such as developer-efficiency at removing defects before testing, developer-efficiency at inserting defects, defects, defects found per person, and the value of defects found. After that, an initial estimate is made to calculate the number of testers based upon the ratio of the baseline project. The initial estimate is adjusted using professional experience to show how the afore-mentioned parameters affect the current project and the baseline project.

3. Project Staff Ratio Method

Project-staff ratio method makes use of historical metrics by calculating the percentage of testing personnel from the overall allocated resources planned for the project. The percentage of a test team size may differ according to the type of project.

4. Test Procedure Method

This model is based on the number of test procedures planned. The number of test procedures decides the number of testers required and the testing time. Thus, the baseline of estimation here is the quantity of test procedures. However, you have to do some preparation before actually estimating the resources. It includes developing a historical record of projects including the data related to size (e.g., number of function points and number of test procedures used) and test effort measured in terms of person-hours. Based on the estimates of historical development size, the number of test procedures required for the new project is estimated.

The only thing to be careful about in this model is that the projects to be compared should be similar in terms of nature, technology, required expertise, and problems solved.

5. Task Planning Method

In this model, the baseline for estimation is the historical records of the number of person-hours expended to perform testing tasks. The historical records collect data related to the work break down structure and the time required for each task so that the records match the testing tasks.