AGENT-BASED PERFORMANCE EVALUATION SYSTEM

ABSTRACT

The problem of performance evaluation is a hydra-headed monster that affects business growth because of non-existence of evaluation process in many organizations. In organizations with an evaluation process in place, performance evaluation is not properly automated; they are usually inconclusive and incomplete due to biased, partial and inadequate handling of performance metrics. In order to achieve the goal of objectivity, credibility and trust in staff appraisal, a system that achieves unbiased, comprehensive and self-information gathering should be provided. This study, therefore, focused on developing an Agent-Based Performance Evaluation System towards addressing these deficiencies.
In developing the proposed system, the spiral model of the Software Development Life Cycle was adopted. The metrics that were used by the system to evaluate performance included attendance, staff’s responsiveness, punctuality and projects. The programming environment used for the development include Netbeans IDE, Hypertext Mark-up Language, MySQL, MySQL DB connector, Apache and PHPmyAdmin. The system/application was evaluated in a social context at T&K printing solutions, Lagos. The company via the Human Resource Manager utilized, tested and evaluated the system on four (4) employees. The evaluation was based on the system’s efficiency and user friendliness/usability.
A model of agent-based performance evaluation system for managing staff-work information and auto-Scoring based on organization’s predefined uniform company-wide appraisal score-weights was developed. An agent-based performance evaluation application was created as an instantiation of the model. The evaluation of the system revealed the manager’s appreciation for the system because employees were glad to have their strengths and weaknesses revealed to them as regards execution of particular projects.
In conclusion, the research found that a proper appraisal of employees can serve as a motivating factor encouraging commitment towards the growth of any organization. It is, therefore, recommended that prompt and unbiased performance appraisal system be deployed by organizations.

CHAPTER ONE/INTRODUCTION 

1.1  Background to the Study  

Largely, the need for evaluation is to reward high performing employees, recognize areas for improvement of staff as well as organizations, and provide recommendation in case of un-reconcilable negligence of staff (Shamsuzzoha, 2013) Though these reasons should make any organization want to adopt evaluation system, most organizations that perform evaluations disrupt the system with too many human interpretations or bias. Others run away from implementing these performance systems due to cost Performance evaluations are usually not as effective as they should be because some organization’s performance process lack credibility; no clarity in the aspect of job being evaluated, no standards against which performance is measured (Cintron, 2011). Performance Evaluation is bias, manual, vague and incomplete in most Organizations (Daoanis, 2012). Most Organizations do not have intelligent systems to appraise their staff thereby adopting manual process which injects subjectivity into staff appraisal (Kateřina, 2013). A recent investigation of major U.S organizations revealed that 40% of managers admitted to forging or controlling performance data because it was clear to them that the evaluation served no valuable purpose in the way it was managed in their Organization (Gordon, 2016).
Appraisal system should track targets, accomplishments and projects of each employee, otherwise, a staff is appraised based on human judgments and traits. Employee is then evaluated by finding a score that best characterizes his or her level of performance for every quality rather than the competence and accomplishment of tasks. A need for automation of the process is indicated. Employee performance is essentially identified with employment obligations, which are expected from a worker, and how well those obligations were accomplished. Managers assess their employees’ performance on a yearly or quarterly basis, keeping in mind the end goal, to help them recognize and recommend areas for improvement (Shaout et al., 2014). The cogent components of an effective appraisal and performance system should include regular feedbacks aimed at providing employees with feedbacks expounding employee’s strengths and improvement areas, to deduce appropriate support plans and to help in decision making on issues such as compensations and recommendations (Chris, 2011). Thus, performance evaluation (PE) is key in enhancing the quality of work input, in inspiring staff and making them more engaged. PE additionally introduces a foundation for increment in monthly remuneration and guides an organization in the development of its employee succession and promotion plans (Shaout et al., 2014).
Several appraisal methods, such as graphic rating scale method, forced choice distribution method, behavioural check list method and so on, were used for employee performance appraisal. New methods like Management by Objectives (MBO) and assessment centres have been suggested (Shaout et al., 2014). The need for an efficient performance evaluation system cannot be overemphasized. Review of existing performance evaluation frameworks and systems revealed subjectivity concerns as performance scoring still remains based on human judgementImpraise (Daoanis, 2012), talentpeak (Cintron, 2011), Performly (Lawler, 2012), PeopleGoal (Katerina, 2013), Namely (Boachie-Mensah, 2012), Tessello (Bretz, 1992), and Fego (Kondrasuk, 2011), ClearCompany (Mayhew, 2016), BambooHR (Capterra, 2015) among others, computer-based applications designed to handle employee appraisals are still subject to human biases because the employer/manager have to play some roles in the appraisal process; giving the score, checking the feedback report, sending the feedback report to be sent to the employee, and so on. The need to rid biasness and quit robbing employees of their promotion calls for an intelligent based performance evaluation system.
 
1.2 Statement of the Problem
The problem of performance evaluation is a hydra-headed monster which includes non-existence of the process in many organizations.  Organizations that put checks and balances into their operations are few. Those who do, spend hours of valuable man-power trying to manually arrive at the staff performance evaluation. When these are done, various data are exposed to personal bias. At other times the performance evaluation are inconclusive and incomplete as there is no clarity of performance metrics. In order to achieve the goal of objectivity, credibility and trust in staff appraisal, a system that achieves unbiased, comprehensive and self-information gathering should be provided.
Computer-based performance evaluation systems in literature have been limited by lack of sufficient automation and lack of consideration of  some relevant evaluation metrics. The search for ideal performance evaluation system remains elusive making it a viable research area.
There is, therefore, the need for a computer-based performance evaluation system that will be staff-centric, task-centric and environment-centric, which this study will focus upon.

1.3 Objective of the Study

The main objective of this study is to develop an agent-based Perfomance evaluation model.
The specific objectives are to:

1. propose a model of agent-based performance evaluation system for managing staff-work information and auto-scoring based on organization’s predefined uniform company-wide appraisal score-weights;
2. develop a prototype of the proposed model, and
3. evaluate the prototype.

1.4 Methodology

The tasks undertaken in order to accomplish this research are as follows:

1. Extensive literature review on the subject matter was carried out.
2. A model for agent-based performance evaluation system for managing staff-work information and auto-Scoring, based on organization’s predefined uniform company-wide appraisal score-weights was developed.

• An application, based on the proposed model, was developed to instantiate the model following the Spiral Software Development Life Cycle. The metrics that were used by the system to evaluate performance include attendance, staff’s responsiveness, punctuality and projects. Netbeans IDE, Hypertext Markup language (HTML), MySQL, MySQL DB connector, Apache and PHPmyAdmin constitute the programming environment.

1. The application was evaluated using an organization as a Case-Study.

1.5 Scope of the Study

The proposed system only targets employee appraisal and does not extend to other areas of the employee career that are not directly linked to the employees’ performance evaluation. The variables that were of interest in the study, therefore, are those that clearly define the employee in terms of performance.
 

1.6 Significance of the Study

As the world innovates in technology, computing assumes everywhere and anywhere paradigm, enabling improved service delivery, and supporting seamless productivity. With new organizations joining the ecosystem of products and services management, it is imperative and highly significant to automate staff performance using unbiased and complete model. Such performance approach must put flexibility in centre stage, thereby allowing organizations achieve transparent performance rating for members of staff.
The proposed approach would enable automated appraisal scores for every staff. A manager or project team head can provide weight values to agreed appraisal metrics. Management can detect staff’s attitude to work automatically through intelligent approaches such as Staff-Computer Inactivity Time and Staff-Email response time, among others. An appraisal administrator can view non-editable appraisal scores at will. The system also allows management provide informed judgements and decisions based on the appraisal outputs.
This approach would go a long way in enhancing work-experiences, staff outputs and overall company achievement as they can share and learn from genuine and unbiased appraisal reportage.