DefinePK

Research Objective

To enhance the performance of an Integrated Model of 360-degree feedback for Administrative Staff in Higher Education Institutions (HEIs) by employing a time series approach to analyze historical data and inform future strategic decisions.

Methodology

The study employed the Autoregressive Integrated Moving Average (ARIMA) model for time series forecasting. Data from July 1, 2020, to December 12, 2021, was collected from administrative staff at the University of Education, Winneba, Ghana. The methodology involved three stages: Identification (determining p, D, and q values), Estimation (using coefficient approximations and model evaluation metrics like AIC and Hannan-Quinn), and Diagnostics and Forecasting (fine-tuning the model based on criteria like BIC, R-squared, and residual analysis). Log-transformation was applied to the data to ensure normal distribution and stabilize variance. Questionnaire validation was performed using Cronbach's Alpha and Kaiser-Meyer-Olkin (KMO) statistics.

Methodology Flowchart

                        graph TD
    A["Data Collection July 2020 - Dec 2021"] --> B["Data Preprocessing: Log Transformation"];
    B --> C["Stationarity Test"ADF""];
    C -- Stationary --> D["ARIMA Model Identification p,d,q"];
    C -- Non-Stationary --> E["Differencing"];
    E --> D;
    D --> F["ARIMA Model Estimation"];
    F --> G["Model Diagnostics Residual Analysis"];
    G --> H["Model Selection"];
    H -- Best Model --> I["Forecasting"];
    H -- Poor Model --> D;
    I --> J["Results and Discussion"];
    J --> K["Conclusion"];                    

Discussion

The research highlights the increasing competition in the higher education sector and the necessity for HEIs to adopt business models, including performance management principles. The study emphasizes the importance of 360-degree feedback as a comprehensive method for evaluating employee performance, offering a more holistic view than traditional single-rated feedback. The application of ARIMA models demonstrates their effectiveness in short-term forecasting of key performance indicators within the Integrated Model of 360-Degree Feedback, suggesting potential for improved employee productivity and strategic decision-making. The study also discusses the challenges of performance appraisal systems and the benefits of a structured, data-driven approach.

Key Findings

- ARIMA models demonstrated high forecasting accuracy for job performance, appraisal satisfaction, and job capability, with low Root Mean Square Errors (RMSE): 0.12 for job performance, 0.04 for appraisal satisfaction, and 0.05 for job capability.
- The ARIMA (1,0,1) model was identified as the best for predicting job performance.
- The ARIMA (0,1,3) model was identified as the best for predicting appraisal satisfaction.
- The ARIMA (4,0,0) model was identified as the best for predicting job capability.
- The study established autocorrelations between appraisal satisfaction and job capability in predicting job performance.

Conclusion

The Integrated Model of 360-Degree Feedback, when analyzed using ARIMA forecasting, proves to be an effective tool for performance appraisal of administrative staff in Ghanaian HEIs. The study confirms that ARIMA models can accurately forecast job performance, appraisal satisfaction, and job capability in the short term, contributing to enhanced decision-making and employee productivity. The research provides theoretical contributions by empirically investigating the link between appraisal satisfaction, job capability, and job performance, and by detailing the ARIMA modeling process for these indicators.

Fact Check

- The study used historical data from July 1, 2020, to December 12, 2021, totaling 76 observations for each variable.
- The ARIMA (1,0,1) model for job performance achieved an R-squared of 0.7942, indicating that 79.42% of the variation in the log of job performance is explained by the model.
- The Kaiser-Meyer-Olkin (KMO) value for questionnaire sampling adequacy was 0.706, which surpasses the suggested minimum threshold of 0.5.

Mind Map