Purpose

This paper attempts to explore possible contributing factors of females' outperformance and males' underperformance in the higher education in Pakistan from teachers' perspective. The central question of the study is what are the key factors that affect female and male students' educational performance at the university level? Using Artificial Neural Network (ANN) as a framework, we attempted to predict differentials of the perceived “female outperformance” and “male underperformance” in higher education. We carried out the study by employing quantitative research methods.

Design/methodology/approach

The data for the study come from 253 teachers from University of the Punjab-largest and oldest University in Pakistan. We used a structured questionnaire for data collection. The analysis was carried out with the help of ANN model. Statistical Package for Social Sciences (SPSS) was used to analyze the data.

Findings

The testing results of ANN indicated 85.3% of teachers' perception was correctly predicted on various dimensions of performance differentials across female and male students in higher education.

Research limitations/implications

The study banks on primary data collected from teachers of the University of University of the Punjab, Pakistan. Thus, the study's universe was limited to one university – University of Punjab. It is purely based on a quantitative approach employing ANN.

Practical implications

The findings of this study have several significant implications, i.e. it makes a significant contribution to the existing body of scholarly texts on the issue of gender reverse change in academic performance in higher education.

Originality/value

The findings of this research, derived from primary data in Pakistan context, qualify this research as an original one. We also claim that this study is one of the first studies on gender reverse change in academic performance among graduate students in a public sector university of Pakistan employing ANN.

中文翻译：

---

教师对高等教育中女学生表现不佳和男学生表现不佳的因素的看法

目的

本文试图从教师的角度探讨巴基斯坦高等教育中女性表现不佳和男性表现不佳的可能影响因素。该研究的中心问题是，在大学层面上影响男女学生教育表现的关键因素是什么？我们使用人工神经网络（ANN）作为框架，试图预测高等教育中“女性表现不佳”和“男性表现不佳”的差异。我们采用定量研究方法进行了研究。

设计/方法/方法

这项研究的数据来自旁遮普大学（Punjab），是巴基斯坦最大和最古老的大学的253名教师。我们使用结构化调查表进行数据收集。该分析是在ANN模型的帮助下进行的。使用社会科学统计软件包（SPSS）来分析数据。

发现

ANN的测试结果表明，在高等教育中男女生表现差异的各个维度上，正确预测了85.3％的教师知觉。

研究局限/意义

研究银行提供的资料是从巴基斯坦旁遮普大学的老师那里收集的主要数据。因此，研究的范围仅限于一所大学-旁遮普大学。它完全基于采用ANN的定量方法。

实际影响

这项研究的发现具有几个重要的意义，即，它对高等教育学业中性别逆向变化问题的现有学术文献做出了重大贡献。

创意/价值

这项研究的发现是从巴基斯坦的主要数据中得出的，因而使该研究具有原创性。我们还声称，这项研究是巴基斯坦公立大学聘用ANN的研究生中在学习成绩上出现性别逆向变化的首批研究之一。