基於勝算比探討學生背景與學習型態對就業表現的影響;Exploring the Effect of Student’s Background and Learning Styles on Employment Performance Based on Odds Ratio

NCUIR > College of Electrical Engineering & Computer Science > Graduate Institute of Computer Science and Information Engineering > Electronic Thesis & Dissertation > Item 987654321/93027

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/93027

Title:	基於勝算比探討學生背景與學習型態對就業表現的影響;Exploring the Effect of Student’s Background and Learning Styles on Employment Performance Based on Odds Ratio
Authors:	林亭延;Lin, Ting-Yan
Contributors:	資訊工程學系
Keywords:	校務研究;學習型態;因果勝算比探勘;勝算比;卡方分析;路徑分析;institutional research;learning style;causal odds ratio mining;odds ratio;chi-square test;path analysis
Date:	2023-07-10
Issue Date:	2024-09-19 16:38:58 (UTC+8)
Publisher:	國立中央大學
Abstract:	本研究旨在探討學生背景與學習型態對就業表現的關係，資料樣本以國立中央大學108年畢業生為研究對象。本研究採用實證資料，資料來源為國立中央大學校務資料倉儲，使用資料欄位包含學生背景資料、UCAN職業興趣診斷與畢業後一年調查問卷。並控制該年度的就業環境和薪資水平這兩項因素所造成的影響。在經資料前處理過後，無資料缺失者共217筆。研究方法為先利用因果勝算比探勘找出資料之間可能蘊含的準因果規則，再透過敘述性統計說明樣本分佈情形，再藉由卡方分析驗證準因果規則是否存在統計相關性，最後使用路徑分析驗證準因果規則是否可能存在因果關係。本研究發現藉由統計方法可驗證因果勝算比探勘所得出之部份準因果規則可能存在因果關係。在分析問題中，因果勝算比探勘與統計方法之間存在先天差異，這可能是導致兩者產生不同結果的原因。若將因果勝算比探勘作為探索資料間可能因果關係的第一步，而不是最終結論，並在後續研究中，使用統計驗證或其他方法來進一步探討。這種方法提供了對尚未研究的問題進行探索，或者探索現有問題中是否存在尚未考慮的因素的可能性。;The purpose of this study is to investigate the relationship between student background, learning styles, and employment performance. The data sample consists of graduates of Nat-ional Central University in 2019. This study utilizes empirical data obtained from the univer-sity′s administrative data repository, including student’s background information, UCAN vo-cational interest diagnosis, and a post-graduation survey conducted one year after graduation. The study controls the impact of employment environment and salary level of that year. After data preprocessing, a total of 217 records with no missing data were included. The research methodology involves first using causal odds ratio mining to identify pote-ntial quasi-causal rules within the data. Descriptive statistics are then employed to describe th-e distribute-on of the sample, followed by chi-square analysis for validating the statistical cor-relation of the quasi-causal rules. Finally, path analysis is used to verify whether causal relati-onships may exist among the identified quasi-causal rules. The study found that certain quasi-causal rules obtained through statistical methods can be validated to potentially indicate caus-al relationships. In analyzing the issue, there are inherent differences between causal odds ratio mining and statistical methods, which may account for the divergent results obtained. If causal odds ratio mining is regarded as the first step in exploring possible relationships and causality bet-ween the data, rather than as final conclusions, and if statistical validation or other methods are subsequently employed for further investigation, this approach offers a means to explore unanswered questions or the possibility of unconsidered factors in existing problems.
Appears in Collections:	[Graduate Institute of Computer Science and Information Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	19	View/Open

社群 sharing

Loading...