失業搜尋與在職搜尋有效性(effctiveness)之比較攸關失業搜尋者最佳策 略之決定，因此實有必要針對「搜尋有效性」課題深入研究，本文之目的即在利 用「大專畢業青年就業狀況調查」資料，分別以樣本統計分析與probit模式估計 尋獲新工作之機率，比較台灣大專青年失業搜尋與在職搜尋之有效性。樣本之統 計分析結果顯示，失業搜尋者顯著地比在職搜尋者使用較多的方法搜尋工作，而 在有關搜尋結果之比較方面則發現，1986年樣本顯示失業搜尋者尋獲新工作之比 例較在職搜尋者為高，而1991年樣本卻顯示，在職搜尋者尋獲新工作之比例略高 於失業搜尋者。利用probit模式估計尋獲新工作機率之實證結果隱含，就1986年 樣本而言，不論是否加入搜尋方法與保留工資因素之考慮，失業搜尋者尋獲新工 作之機率均較在職搜尋者為高。1991年樣本之實證結果與樣本統計分析結果之涵 義則不盡相同，若不列入搜尋方法與保留工資因素之考慮，尋獲新工作機率之 probit實證結果顯示，失業搜尋者尋獲新工作之機率與在職搜尋者並無顯著差 異，但若加入搜尋方法與保留工資因素之考慮，則失業搜尋者尋獲新工作之機率 較在職搜尋者為高。

　　　　　The relative effectiveness of employed versus unemployed searchis an important issue addressed in the empirical job search literature; the answer to this question is critical to tlie validity of theory of search unemployment. In this paper we apply descriptiveanalysis and probit analysis to examine the relative effectivenessof employed and unemployed search for 1984/85 and 1991 collegegraduated youth in Taiwan. For 1984/85 college graduates, results from both the descriptive analysis and probit analysis suggestthat unemployed job seekers are more effective in search by gaining more new employment. The differences in search methods usedbetween employed and unemployed searcliers may explain some ofthese outcomes. For 1991 college greduates, the descriptive analysis result indicates a higher job-finding rate for employed searcliers.However the probit esimation results imply that the probability ofgaining new employment is higher for unemployed searchers whenthe search choice variables are included.