WWW.THESIS.XLIBX.INFO
FREE ELECTRONIC LIBRARY - Thesis, documentation, books
 
<< HOME
CONTACTS



Pages:     | 1 |   ...   | 4 | 5 || 7 | 8 |   ...   | 9 |

«ABSTRACT The purpose of this research study is to perform a comparison and contrast analysis on three F. Davis Cardwell published longitudinal case ...»

-- [ Page 6 ] --

One concern to be aware of in reference to the KLOEC metric is that programming languages have different levels of instruction explosion. Currently, digital computers only “understand” machine languages based on the ones and zeroes of binary. Programming languages are used to translate between the natural languages of humans and machine languages of computers; the problem is that different programming languages translate between human languages and machine languages with differing levels of efficiency. For example, code for a database lookup may take 20 lines of executable code when written in the programming language “C” whereas the same problem can be solved by using a five line statement in the database query language SQL. (Burd, 2010, pp. 273 – 275) Regardless, KLOEC is a convenient metric for comparing software development projects because it is readily available; it is often provided by automated testing tools.

Additional methods used for calculating productivity in the cases studied include user stories/personmonth, function points/person-month, and the Putnam Productivity Parameter. Both user stories and function points attempt to uncouple the measurement from any peculiarities of the programming language used by focusing on the implemented functionality. User stories are an agile approach for capturing requirements using informal communication. Each discrete function needed by a system has an associated user story. (Ho, et al., 2006, p. 3) The function points method categorizes each discrete function and assigns a weight to it based upon past development performance. The cost of one function point is then calculated, and the results used to manage the development effort. (Boehm, Abts, Brown, Chulani, Clark, Horowitz,... Steece, 2000, pp. 15 – 17) The Putnam Performance Parameter is used to normalize differences in the software development environment, based on observation of a dozen large software development projects. [See Appendix D for more information on the Putnam Productivity Parameter.] Some additional areas that can impact development costs are fault density and ergonomic factors.

Fault density is measured in terms of total defects [TD] divided by thousands of lines of executable code [KLOEC.] Defects represent wasted effort; a process that decreases fault density reduces development costs. Ergonomic factors relate to the development environment. In regard to the cases studied, XP has an inherent advantage compared to traditional methods. The preference for XP labs is an open development lab, compared to the private or semi-private cubicles preferred by traditional development; the use of an open development lab can result in lower facilities costs due to the more efficient use of floor space. Likewise, the emphasis on pair programming can result in lower capital equipment costs, since each team member does not require a dedicated desktop computer.

The Sabre-A study experienced an increase in productivity and a decrease in fault density when XP methods replaced traditional methods. (Layman, et al., 2004a., p. 8) The Sabre-P study experienced a decrease in fault density when XP methods were used, compared to published industry averages.

Productivity was measured using two different methods: KLOEC/person-month and function points/person month. Using KLOEC/person-month, productivity measured using XP methods was similar to published industry averages. Using function points/person-month, productivity measured using XP methods was higher than published industry averages. (Layman, et al., 2004b., p. 8) The IBM study experienced an increase in productivity using all three measurement methods, and a decrease in fault density when XP methods replaced traditional methods. (Williams, Krebs, et al., 2004, p. 8) In light of these results, the null-hypothesis that there are no significant differences between development costs of traditionally developed software projects and projects developed using agile methods is not supported by the evidence presented.

THE PERFORMANCE OF AGILE METHODS:COMPARISON TO TRADITIONAL DEVELOPMENT METHODS

SCHEDULE COMPARISONS

Schedule performance refers to how well a project progresses compared to the original estimate of the amount of time needed to finish the project. It is important to remember that the original schedule is by necessity an estimate—an educated guess as to the amount of time that the development teams believes it needs to complete the planned scope of work. Project schedule performance is one of the factors that differentiate between a failed project and a successful project. (Linberg, 1999, p. 1) Both productivity and quality can impact schedule performance. When considering productivity and schedule performance, it is important not to fall into the trap of assuming that increased productivity automatically correlates with improved schedule performance. While it is attractive to assume that if a development method allows more output for a given amount of input, the method would also support the originally planned amount of output for a given amount of input in an accelerated period. This assumption does not jibe with practical experience; for one thing, it fails to take into account that one or more critical paths exist in every development project. A critical path refers to the longest path through the project schedule; it determines the earliest possible time that the project can be brought to completion. Another characteristic of a critical path is that a critical path has the least amount of float or slack time, which represents “wiggle room” in the schedule, i.e., the amount of time that an activity can be delayed without delaying another activity or the project’s scheduled completion date.





(Schwalbe, 2011, p. 228) Understanding of the critical path allows for effective schedule management;

if activities on the critical path can be accelerated, it enables the possibility of compressing the project schedule. Two methods used to accelerate projects are crashing and fast tracking. Crashing refers to adding resources to a task in order to complete it faster than scheduled, while fast tracking refers to performing tasks in parallel that were originally scheduled to be performed in series. While both methods can be used to improve schedule performance, they both have associated drawbacks.

Crashing often results in increased project costs due to the addition of resources. Fast tracking can lead to wasted effort and rework when activities are completed out-of-sequence. The critical path does not illustrate the shortest path through the project, nor does it illuminate the critical tasks of the project.

Multiple critical paths can exist through a project, and the critical path can change as the schedule changes in response to completed project activities and project schedule challenges. The critical path displays the best-case time estimate through a project, essentially saying: This the best possible time we can make from the beginning of the project to the end, using the schedule as it now exists.

(Schwalbe, 2011, p. 233) Another reason that productivity gains do not automatically translate to improved schedule performance is that project dependencies can have a significant impact on schedule performance. A dependency refers to the sequencing of project activities within the project. There are three types of

project dependencies:

1. Mandatory dependencies, also known as hard logic, are innate features of the work to be performed. For example, the ability to perform the testing-the-code activity depends on the completion of the writing-the-code activity.

2. Discretionary dependencies, also known as soft logic, are dependencies that are defined at the discretion of the project team. For example, it is good practice to refrain from starting the final user interface design until the customer has an opportunity to evaluate the prototype and provide feedback to the development team.

3. External dependencies are caused by interaction between non-project and project activities. An example is that the installation of a developed software package may depend on the installation of a new operating system onto the existing hardware by an external contractor. Even though the installation of the operating system may be outside of the scope of the project, this creates an external dependency because late installation of the operating system will adversely affect the project schedule. (Schwalbe, 2011, p. 217)

THE PERFORMANCE OF AGILE METHODS:COMPARISON TO TRADITIONAL DEVELOPMENT METHODS

While productivity can affect the project schedule, it is important to understand its limitations. Recall from the Costs section that productivity is usually calculated using the formula KLOEC/PM.

Productivity changes affecting costs usually affect the numerator of the productivity formula, while productivity changes affecting schedule performance usually affect the denominator of the formula.

The denominator of the formula is the person-month. Brooks (1995) points out that reliance on the person-month for measuring effort used on software development projects can be misleading. Implicit in the person-month formula is the idea that people and time are the same, which is simply not supported by software development experience. Software development is an inherently complex activity, relying heavily on communication. If people and time were interchangeable, it would be possible to achieve any level of schedule performance desired simply by adding more people. In the original example, assume that the project which originally required two months for completion was instead needed in one week. If people and time were equivalent, the project manager in this case could simply assign a total of 36 people to the project; the original person-month metric of six is preserved, with the decrease in time to completion offset by the additional development staff. One factor that this scenario fails to incorporate is that communication is critically important to creative processes such as software development, and that adding more people to a project increases the number of communication channels. (Schwalbe, 2011, p. 395) The number of communication channels is calculated by the formula n(n-1) / 2 where n is equal to the number of people involved in the project. In the original example, there is one communication channel. In the accelerated project, there are 630 communication channels. Brooks notes that adding people to a software development project can actually decrease team productivity. (Brooks, 1995, pp. 6 – 10) Nonetheless, the personmonth is a convenient metric for comparing software development projects, because it is readily available for any managed project.

Project quality has a clearer relationship to schedule performance than productivity does. Quality for the cases studied is measured as a function of defects divided by thousands of lines of executable code [KLOEC.] Further, quality is divided into two categories, internally visible quality and externally visible quality. Internally visible quality refers to defects that are found and corrected during development, while externally visible quality refers to defects that are found after the product is released to the customer. Generally, increased defects negatively affect the project schedule, while decreased defects positively affect the project schedule. Quite simply, correcting software defects takes time; the more defects existing in the software, the longer it takes to correct them. (Pfleeger & Atlee, 2010, p. 426) An additional area that can impact schedule performance is the method by which the software is delivered to the customer. Traditional software development projects usually gather requirements from the customer, develop the software based on the requirements, and then deliver the software to the customer once it’s complete. Agile development projects, on the other hand, usually stress continual delivery of the product, even if the product isn’t complete and provides only partial functionality.

Additionally, there is usually a great deal more interaction with the customer throughout the development project when agile methods are used, compared to traditional methods. This additional interaction can result in improved delivery of the desired functionality; increased customer interaction and continual product delivery allow the team to catch and correct mistakes made during the requirements gathering phase, thus ensuring that the product delivered solves the problem that it was intended to solve. Developers of traditional projects, on the other hand, may not be aware that they are solving the wrong problem until the product is delivered to an unhappy and unsatisfied customer.

(Wells, 2009) Although this is a clear advantage of agile methods, it does tend to confound the comparison of agile and traditional schedule performance.

The Sabre-A study experienced an increase in productivity and a decrease in fault density when XP methods replaced traditional methods. (Layman, et al., 2004a., p. 8) The Sabre-P study experienced a decrease in fault density when XP methods were used, compared to published industry averages,

THE PERFORMANCE OF AGILE METHODS:COMPARISON TO TRADITIONAL DEVELOPMENT METHODS

while productivity was similar to or higher than published industry averages. (Layman, et al., 2004b., p. 8) The IBM study experienced an increase in productivity using all three measurement methods, and a decrease in fault density when XP methods replaced traditional methods. (Williams, Krebs, et al., 2004, p. 8) Recall from the costs section that costs, schedule and quality are interrelated; for any project, an alteration of two of these facets results in an induced change to the third. (Dillman, 2003, p.



Pages:     | 1 |   ...   | 4 | 5 || 7 | 8 |   ...   | 9 |


Similar works:

«Dosage Calculation Tutorial Kent State University Regional Campuses Karol Burkhart Lindow, RN,C, MSN, CNS Associate Professor, Nursing Reviewers: Denise Ayers, RN, MSN Joan Lappin, RN, MSN Summer 2004 copyright c 2004 Kent State University Contents Acknowledgements 4 1 Introduction 5 1.1 What is dosage calculation?................... 5 1.2 Doesn’t the pharmacist do all this?............... 6 1.3 I thought the doctor was the one responsible for medication...»

«The Forged Note A Romance The Forged Note: A Romance of the Darker Races (1915) Of The Darker Races 1915 Otherwise, Manage services have now based on the shows on they may add off. The monthly proficiency will vacate you up contracts what give too however little brochures. Want all your services good of you use those feeling. Of hearing to services that will pass to meet a key strategy if a property for the translation will dictate the important plan means will sustain more big to name purchase...»

«Job Aid: First American Comprehensive Calculator (FACC)-TRID Quoting The purpose of this document is to provide general instructions on how to use the First American Comprehensive Calculator (FACC) for a TRID-related Quote for a Purchase and a Refinance. Table of Contents Sale with a Mortgage: Accessing the FACC Recording Documents Summary Screen Delivery Options...12 Refinance: Entering Transaction Information Entering Title Rates and Endorsements Closing Information Recording Documents...»

«OFFICE OF THE NEW YORK STATE COMPTROLLER Thomas P. DiNapoli • State Comptroller Growing Cracks in the Foundation: Local Governments Still Challenged to Keep Up with Vital Infrastructure Needs DIVISION OF LOCAL GOVERNMENT AND SCHOOL ACCOUNTABILITY FOLLOW-UP REPORT – SEPTEMBER 2014 For additional copies of this report contact: Comptroller’s Press Office 110 State Street, 15th floor Albany, New York 12236 (518) 474-4015 or email us: localgov@osc.state.ny.us www.osc.state.ny.us To be removed...»

«M.R.C. TECH 6 INSTRUCTION ADDENDUM PART NO. 0001200 2-AMP RATING Thank you for purchasing this new Tech 6 Train Control System. This advanced train control system can control two types of locomotives; standard, analog D.C. locomotives and new sound-equipped locomotives that contain N.M.R.A. compatible sound or non-sound decoders. The new Tech 6 line of power packs will take you beyond the 21st Century of train control. Quick Start Simply plug the power supply into your new Tech 6. Connect two...»

«Traders‘ Interview Andrea Unger Der Stoff, aus dem die Helden sind Es gibt gute Trader und es gibt schlechte Trader. Das leuchtet ein. Andererseits wirft dieser Satz sofort die Frage auf, welcher Trader der Beste ist. Und so gibt es im Trading, wie in jedem männlich dominierten Bereich, der auch nur im Entferntesten so etwas wie einen Wettkampfcharakter erkennen lässt, Wettbewerbe, Weltmeisterschaften, Preisverleihungen. Es geht um Ruhm, es geht um Ehre. Im Jahrestakt gibt es neue...»

«.. th holidays: –  –  – rd ® Bot –  –  – –  –  – –  –  – –  –  – –  –  – –  –  – –  –  – –  –  – –  –  – Delivery. –  –  – –  –  – –  –  – –  –  – –  –  – –  –  – informs or make –  –  – –  –  – –  –  – –  –  – –...»

«Steps to Success: An Instructional Design for Early Literacy Mentor-Coaches in Head Start and Early Head Start Unit 1: Building Relationships to Promote Child Literacy Outcomes Early Literacy Module 4 Mentor-Coaching and Supervision MENTOR-COACH MANUAL Unit 1: Building Relationships to Promote Child Literacy Outcomes MODULE 4 EARLY LITERACY MENTOR-COACHING AND SUPERVISION CONTEXT Mentor-coaching is most successful when all of the professionals involved understand and recognize their roles and...»

«Fast Allocation and Deallocation with an Improved Buddy System∗ Gerth Stølting Brodal† BRICS‡ Department of Computer Science, University of Aarhus, IT-Parken, ˚bogade 34, DK-8200 ˚rhus N, Denmark A A gerth@brics.dk Erik D. Demaine§ MIT Computer Science and Artificial Intelligence Laboratory 32 Vassar Street, Cambridge, Massachusetts 02139, U.S.A. edemaine@mit.edu J. Ian Munro School of Computer Science, University of Waterloo Waterloo, Ontario N2L 3G1, Canada imunro@uwaterloo.ca...»

«UNITED STATES OF AMERICA Before the COMMODITY FUTURES TRADING COMMISSION ~-. :· ·-.-:' ~ ) :~ 0 ;~== :::~~ ) CFTC Docket No: 12-30 n f. In the Matter of ). (,) PROCEEDING~ ) ORDER INSTITUTING PURSUANT TO SECTIONS 6(~) AND~) Maksim Baron, ) ) OF THE COMMODITY EXGJIANGE._;. ) ACT, AS AMENDED, MAKING Respondent..J:: ) FINDINGS AND IMPOSING REMEDIAL ) SANCTIONS ) I. The Commodity Futures Trading Commission (Commission) has reason to believe that from March 2010 to at least October 2010 (the...»

«SERIES 198 PREFERENCE SHARE TERMS AND CONDITIONS The following are the terms and conditions (the Conditions) of the Series 198 Index linked redeemable preference shares (the Preference Shares) issued by Eukairos Investments Ltd (the Company) on 24 September 2013. Terms not otherwise defined have the meanings given in Condition 1 (Definitions) below. References to a numbered Condition shall be to such numbered section of the Conditions. In the event of any inconsistency between the Articles and...»

«Standard Operating Procedures Flight Operations Briefing Notes Conducting Effective Briefings Flight Operations Briefing Notes Standard Operating Procedures Conducting Effective Briefings I Introduction To ensure mutual understanding and effective cooperation among crewmembers and with ATC, in-depth takeoff and approach / go-around briefings should be conducted on each flight.A thorough briefing should be conducted regardless of: How familiar the departure airport / SID or destination airport /...»





 
<<  HOME   |    CONTACTS
2016 www.thesis.xlibx.info - Thesis, documentation, books

Materials of this site are available for review, all rights belong to their respective owners.
If you do not agree with the fact that your material is placed on this site, please, email us, we will within 1-2 business days delete him.