COST ESTIMATION TECHNIQUES AND COCOMO

Cost Estimation Techniques

• 1-)Algorithmic cost modelling• 2-)Expert judgement• 3-)Estimation by analogy• 4)-Parkinson’s Law• 5-)Price to win

Algorithmic Cost Modelling

• Model uses a mathematical formula to predict project costs based on estimates of the project size, the number of engineers and other products.

Expert Judgement

• Several experts estimate a project cost.• These estimated costs are compared and

discussed.• This process iterates until an agreed estimate

is reached.

Estimation by analogy

• Cost estimation is done by looking at the previous completed projects.

Parkinson’s Law

• The cost is determined by available resources.

Price to win

• The estimated effort depends on the customer’s budget and not on the software functionality.

Algorithmic Cost Modelling in Details

• Most general form, this model can be expressed as;

• Effort = A ´ SizeB ´ M• Where A is a constant factor depends on the

organisational practices and the type of software that is developed.

Algorithmic Cost Modelling in Details

• Size is either code size or the points gathered by function and object points.

• The value of exponent B usually lies between 1 and 1.5. Depends on the complexity of the code.

• M is calculated by considering some attributes such as dependability requirments and the experience of the development team.

ACM suffers from fundamental difficulties

• Size is hard to estimate at an early stage when only a specification is available. Function-points and object-points are easier but, still often in accurate.

• The estimations of B and M are subjective.

ACM Info

• If you use the model, you should develop a range of estimates(worse,expected and best) and apply the formula for all of them.

• More information, more the accuracy of the cost estimation will be.

x

2x

4x

0.5x

0.25x

Feasibility Requirements Design Code Delivery

The COCOMO Model

• Is an algorithmic model.• Well documented, available in the public domain.• Has been used widely.• Most recent version is Cocomo 2• Assumes that the software is developed according

to a waterfall process.• In the new version, it recognises prototyping,

component composition and database programming.• Cocomo 2 supports spiral model.

About the First Version

• Three-level model where the levels correspond to the analysis of the cost estimate.

Projectcomplexity

Formula Description

Simple PM = 2.4 (KDSI)1.05 M Well-understood applicationsdeveloped by small teams.

Moderate PM = 3.0 (KDSI)1.12 M More complex projects whereteam members may have limitedexperience of related systems.

Embedded PM = 3.6 (KDSI)1.20 M Complex projects where thesoftware is part of a stronglycoupled complex of hardware,software, regulations andoperational procedures.

Cocomo 2

• There are sub models that are part of the Cocomo 2 such as;

• 1-An application-composition model• 2-An early design model• 3-A reuse model• 4-A post-architecture model

An application-composition model

• Assumes that the systems are created by composing existing components.

• The formula as follows;PM = (NAP*(1- %reuse/100)) / PRODWhere NAP stands for total number of object points. Reuse stands for estimation of the amount of

reused code PROD stands for object-point productivity.

Object-Point Productivity

DeveloperÕs experienceand capability

Very low Low Nominal High Very high

ICASE maturity andcapability

Very low Low Nominal High Very high

PROD (NOP/month) 4 7 13 25 50

An early design model

• Once the user requirments are agreed, without a detailed architectural design, this model can be used to estimate.

• The formula as follows;

PM = A ´ SizeB ´ M whereA stands for the size of the system.B is not fixed as in Cocomo 81, but it varies from 1.1 to 1.24.M stands for 7 project and process characteristics such as;M = PERS ´ RCPX ´ RUSE ´ PDIF ´ PREX ´ FCIL ´ SCED;

An early design model continues

• The list of 7 M project and process characteristics• 1-RCPXreliability and complexity• 2-RUSEreuse required• 3-PDIFplatform difficulty• 4-PERSpersonnel capability• 5-PREX personnel experience• 6-SCEDschedule• 7-FCILsupport facilities

A reuse model

• This model is used to estimate the effort required to integrate reusable or generated code.

• The formula is ESLOC = ASLOC * (1-AT/100) * AAM. where ESLOC stands for new source code to be coded. ASLOC stands for the number of lines code to be

adapted. AT is the percentage of the code to be adapted. AAM is the sum of three components.

A Reuse Model Continues

• AAM = AAF + SU + AA• AAF represents the cost of making changes to

the reused code. To do this, you have to take into account design, code and integration changes.

• SU is the costs of understanding the code.• AA is the costs of reuse decision making. It

varies 0 to 8 depending on the amount of the analysis effort required.

A post-architecture model

• This model uses the same formula• PM = A ´ SizeB ´ M• But with more information about the software

to be developed, more attributes are considered, so the estimation becomes more accurate. For example, Size of database used(DATA), Memory constraints(STOR),personnel continuity(PCON)

A post-architecture model continues

• For the exponent B is based on five scale factors. These factors are rated on a six-point scale from very low to Extra high(5 to 0).

• To calculate B;• B =1.01 + (Sum of ratings) / 100

Scale Factors

Precedentedness Reflects the previous experience of the organisation with this type ofproject. Very low means no previous experience, Extra high meansthat the organisation is completely familiar with this applicationdomain.

Developmentflexibility

Reflects the degree of flexibility in the development process. Verylow means a prescribed process is used; Extra high means that theclient only sets general goals.

Architecture/riskresolution

Reflects the extent of risk analysis carried out. Very low means littleanalysis, Extra high means a complete a thorough risk analysis.

Team cohesion Reflects how well the development team know each other and worktogether. Very low means very difficult interactions, Extra highmeans an integrated and effective team with no communicationproblems.

Process maturity Reflects the process maturity of the organisation. The computationof this value depends on the CMM Maturity Questionnaire but anestimate can be achieved by subtracting the CMM process maturitylevel from 5.

Calculate the ratings

• Precedentedness - new project for the organisation.- rated low (4)

• Development flexibility - no client involvement - rated very high (1)

• Architecture/risk resolution - No risk analysis – rated very Low .(5)

• Team cohesion - new team - nominal (3)• Process maturity - some control - nominal

(3)

Effects of Cost Drivers

Exponent value 1.17System size (including factors for reuseand requirements volatility)

128, 000 DSI

Initial COCOMO estimate withoutcost drivers

730 person-months

Reliability Very high, multiplier = 1.39Complexity Very high, multiplier = 1.3Memory constraint High, multiplier = 1.21Tool use Low, multiplier = 1.12Schedule Accelerated, multiplier = 1.29Adjusted COCOMO estimate 2306 person-months

Reliability Very low, multiplier = 0.75Complexity Very low, multiplier = 0.75Memory constraint None, multiplier = 1Tool use Very high, multiplier = 0.72Schedule Normal, multiplier = 1Adjusted COCOMO estimate 295 person-months