工作满意度调查研究外文翻译.docx

1、附 件 3外 文 翻 译原 文 1J ob sa tisfa ction and qua lity of wor king life: A reassessmentThe usefulness of job satisfaction m easures in assessing job characteristics and in improving quality of w orking l ife i s problematic. Public opinion polls and org anizational survey s have show n hig h and stable l

2、evels of job satisfaction for many y ears, w hile s ig ns of em ploy ee frustration and a l ienation have been in-creasing . C loser ex amination of this paradox ica l finding leads to the conclusion that, w hatever rig orous job satisfaction survey s are measuring , it i s not the inform a tion nee

3、ded to modify jobs and w ork in order to reduce employ ee frustration. To overcom e flaw s in the measurement of satisfaction for this purpose em ploy ees themselves need to be more involved in the measurement process. R ecent ex perience w i th em ploy ee participation in action research on quality

4、 of w orking li fe i s noted, and the ex tension and furtherance of this model i s proposed.J ob satisfaction as a concept has become an em barrassing ambig uity. For m any investig a tors in the social sciences, an interest in job satisfaction frequently represents an interest in quality of w orkin

5、g l i fe and industrial hum anism, and sug g ests a concern w i th improving the ex perience of people w ith jobs and w ork. S ince the 1930 s this interest has been concerned w i th monitoring the factory model of w ork desig n w hich has been diffused from manufacturing into the service and cleric

6、al sectors. It must be asserted, how ever, that m uch of w hat has passed for job satisfaction research has fa i led to study the job or the w ork i tself. J ob satisfaction research has historically been used e i ther to support or to a ttack the sta tus quo, and the trend continues.The embarrassme

7、nt w i th respect to job sa tisfaction m easurement i s that survey s of A merican employ ees continue to show that ex tremely hig h percentag es of those measured report satisfaction w i th their jobs, w hile a t the same time the incidence of decreased w orker com m i tment as expressed throug h i

8、ncreases in absenteeism ( especially part-w eek absenteeism), strikes ( for other reasons than w ag es), employ ee rejection of neg otiated contracts, and sabotag e of product and plant, is hig h and apparently becoming g reater. Increasing public a ttention has been draw n tothese i ssues of employ

9、 ee a l ienation. If public concern continues to g row for w hat i s increasing ly seen as employ ees quiet desperation w i th w ork, then the pressures for a reconciliation betw een these facts and the carefully prepared, rig orous job satisfaction survey s w i l l become inex orable.A IMThe presen

10、t paper i s intended to address this reconciliation not in terms of reducing sam pling errors, g uarding against response set, or the use of m ore sophisticated statistical tests of satisfaction data as currently defined and measured, but in terms of aquite different mode of research. In so doing I

11、do not seek to criticize the job satisfaction research per se, but ra ther to criticize the use of that variable in evaluating and a ttem pting to improve the quality of w orking l ife . The particular research model proposed i s an action research approach ( L ew in, 1946 ; Davis, 1971 ) model of d

12、em ocratization of job desig n ( Elden, 1976 ; Tay lor, 1976 ; Herbst & Getz, 1977 ) w hich indicators of what w ould improve the quality of w orking li fe for em ploy ees in g iven w ork setting s a re applied w ith the involvem ent and comm itment, and in the lang uag e , of those concerned. In th

13、is m odel i s captured the ability to measure new criteria more relevant to the personal development of the job holder, as recently described by B lackler & B row n ( 1975 ) .A P P R OA C HThe i ssue of w hat the job satisfaction statistics really m ean has been previously approached from several po

14、ints of view. The debate over unitary versus multiple measures of satisfaction w i th w ork has long interested investig a tors, beg inning w i th the intrinsic versus ex trinsic factors studied by Kahn & M orse ( 1951 ) , and finding a forum most recently in W ork in A merica ( 1973 ) . The present

15、 position in that debateurg es the use of new, specific measures of various job characteristics, increasing l y projective measures using hy pothetical cases, or otherw ise approaching the respondent more cautiously or w i th indirect questions because direct questions ma y be too threatening . This

16、 debate, a lthoug h heated, rem a ins la rg e l y unresolved, in spite of rig orous and carefully done studies such as the S urvey of W orking C onditions, 1970 from the U niversity of M ichig ans Institute for S ocial R esearch ( IS R ) w hich undertook to compare the unitary and additive approache

17、s to job satisfaction scales.In this IS R survey, J OB S AT 7 0 ( the additive measure) and Overa l l J ob S a ti sfaction ( the unitary measure) w ere found to be related to each other a t surprising l y low levels ( Herrick& Quinn, 1971 ) .2S ome other of the recent job satisfaction discussions to

18、ta l ly a void e ither definitional or measurem ent debates in favour of more direct appeals to the underly ing social i ssues. For ex ample, part of the support for industrial hum anism has been g enerated by ex trapolating the absolute numbers of A m erican w orkers presumed reporting dissatisfact

19、ion based on the job satisfaction statistics currently ava ilable ( R osow, 1974 ) . U sing this doug hnut vs. hole approach, dissatisfaction i s defined as an im portant social problem because the w orkers w ho report displeasure w i th their w ork must number in the millions.T H E P R OB L E MIn s

20、pite of these reconceptualizations the supreme authority for the state of A merican w orkers still seems to be the percentag es from job satisfaction indices. W hether these indices are the crude s ing le i tem measures taken in Gallup or other national polls, or are the very sophisticated multiple

21、i tem scales such as those a lready m entioned, the proportion of w orkers reporting satisfaction rem a ins inex plicably hig h. The statistical fact i s that, reg a rdless of w hat deg ree of measurement sophistication is broug ht to bear, 80 per cent or more of those A merica ns survey ed report b

22、eing satisfied w ith their jobs. This i s true w hether the studies use data specific to w orkers in assembly plants or to national random samples.A Imberman, of the consulting firm of Imberman and De Forest of C hicag o, reported a survey of 3800 employ ees in five factories w hich revealed that 79

23、 -85 percent reported satisfaction w ith assem bly l ine w ork ( Im berman, 1972 ) . R esearchers a t the R utg ers U niversity M edical S chool reported that of 576 U AW members interview ed in 1968 , 95 per cent w ere satisfied w i th their jobs in an auto plant ( S iassi,C rocetti & S piro, 1974

24、) . A l thoug h these investig a tors sta te that their sample w as representative of an insured g roup of U AW members in a prepaid union health plan in B a l timore, i t should be noted that their sample i s characterized as w hite males, avera g ing 40 y ears of ag e, about 13 y ears service, and

25、 earning $ 9000 or more annually.These results a re consistent w ith more carefully sa m pled national survey s reported from time to time. For exa mple, a 1954 national survey of half a million w orkers by S c ience R esearch A ssociates ( S R A ) of C hicag o reported 81 per cent of those polled w

26、 ere satisfied w i th their w ork. M ore recently the Gallup org anization has reported 87 percent satisfied in a 1964 poll, and 77 per cent satisfied in 1973 . Very recent survey results, reported by the S urvey R esearch C enter a t IS R , reveal fully 91 per cent of m a le w orkers are satisfied

27、w i th their jobs. These l ast results are as hig has those reported in earlier survey s ( w hich a re sum marized over the period 1958 -1973 by Kaplan, 1976 ) .These different studies a l l clearly sug g est that an overw helm ing majority of A m erican w orkers report satisfaction w i th their w o

28、rk. These results a l so show l i ttle chang e ( only four percentag e points) over the 20 -y ear interval betw een the boom y ears 1954 -1973 . It seems that under the rang e of most norm a l circumstances job satisfaction ( or the absence of dissatisfaction) rang es from a low of about 79 percent

29、to a hig h of 95 percent. This trend is apparently not ex clusively an A m erican phenomenon, as C herns ( 1975 ) points out in a discussion of recent job satisfaction research in the U nited King dom.This pattern receives additional support from the exam ination of data sy stema tically collected f

30、rom som e 20 000 employ ees a t a l l levels ( non-supervisory to manag ement) in a variety of different org anizations. The C enter for R esearch on the U ti l ization of S cientific Know ledg e ( C R U S K) a t the U niversity of M ichig an collected these data betw een 1966 and 1970 from some 33

31、offices and plants in 15 U S companies( C R U S K, 1970 ; Tay lor & B ow ers, 1972 ) . A lthoug h they w ere not sy stematically sampled, these org anizations differed w idely on dim ensions l ike manag ement philosophy and economic condition; as w ell as s ize, technolog y, collar colour and the l

32、ike. Overall, 85 percent of a l l 20 000 people reported being satisfied ( or more specifically not being dissatisfied) w i th their jobs, w hile across these org anizations the rang e of this sta ti stic w as quite narrow. A n insurance office of 200 people topped the l i st of 33 org anizations w

33、i th 95 per cent not dissatisfied. A pa per mill employ ing 440 people set the low point am ong the 33 org anizations w ith 76 per cent reporting no dissatisfaction w ith their jobs. In spite of not being taken from a scientific sample of A merican org anizations, this rang e of satisfaction reporte

34、d i s not unlike the national survey s a lready described.A t the sam e time that w e find this overw helming proportion of employ ees in a ll of these very different org anizations reporting they a re not dissatisfied w ith the w ork, w e find indications that this reliable m easure of job satisfac

35、tion i s not as hig hly related to absenteeism, w i thin those org anizations, as w e mig ht ex pect. In the same studies w i th g rievances and turnover measures ( w here less data are available), the relationships w ith job satisfaction a re not consistently hig h or low. If w e look a t differenc

36、es am ong w ork g roups w i thin these org anizations the causal relationships betw een satisfaction and org anizational behaviour are not very hig h ( Tay lor &4B ow ers, 1972 , pp.77 -79 , 89 ) . These finding s ma y be ex plained on the basis that, reg a rdless of how w e l l w e define w ork sat

37、isfaction and how many careful categ ories w e separate and recom bine i t into, w e a re stil l measuring more than perceptions of the w ork i tself. We cannot expect to measure a ll that i s important to w orkers or to g et bey ond their internal defences or expectations w ith precoded questionnai

38、re measures a lone.Aut hor : J AMES C. TAYLORNa t i ona l i t y : US AOr i g i na t e f r om: J our na l of Oc c upa t i ona l Ps y c hol og y , 1977 . Vol . 50译 文 1重新评估工作满意度和工作生活质量有效的工作满意度的措施在评估工作特性和提高工作生活质量中是有问题 的。民意调查显示，虽然多年以来，工作满意度水平高且稳定，员工的挫折感和 异化迹象却一直在增加。经过更密切的检查，这调查得出的结论是矛盾的，即无 论再严谨的工作满意度调查及测

39、量，得到的只是减少员工的挫折感和修改工作两 方面的没必要的信息。根据以往的经验以及对工作生活质量的研究表明，为了克 服这个缺陷，在测量工作满意度的时候，员工本身需要更多地参与测量。工作满意度已经成为一个模糊不清的尴尬概念。许多代表着工业人文主义利 益的社会科学调查员都对工作满意度十分有兴趣，他们提出了建议：要提高工作 满意度就必须关注和改善人与职位的关系。从 1930 年开始，这种兴趣引起的关 注已经从制造业扩展到服务和文职部门。然而，可以断言，很多过去的工作满意 度的研究都没有研究工作或工作本身。历史上曾经有过对工作满意度的研究，这 种可以支持或者攻击现状的趋势还将继续下去。在对美国员工工作

40、满意度的继续调查研究中，出现的尴尬是他们工作的 满意程度是用极高的百分比来衡量的，而在同一时间内降低对工人的承诺，员工 通过增加缺勤率，尤其是部分缺勤率，而拒绝谈判达成的合同以及破坏产品的比 率显然变得更大，这是除了工资以外原因引起的罢工。员工异化的这些问题已经 提起公众的注意，但是如果公众的关注继续增长，为什么员工安静的工作越来越 被看作是压力的缓解，这些精心的准备和事实将使对工作满意度的严格调查成为 必然。前言 本文是以在不减少防范反应和抽样误差的情况下，使用更精密的统计测试为目的，对满意度的数据在不同的模式下进行界定和衡量。我这样做并不是刻意对 工作满意度研究本身进行批评，而是对使用该变

41、量的评价并试图改善工作生活质 量进行批评。这里要特别提出的一个研究模式是行动研究方法( Le wi n，1946； Da vi s ，1971) ，一个关于职务设计的民主化模型( El den，1976；Ta y l or ，1976； He r bs t & Ge t z，1977) 。那些指标会改善值得关注的工作生活质量和员工在特定 的工作设置中的参与，以及使用的语言。在这个模型中，要抓获具备能力衡量更6多有关个人发展新标准的工作持有人，这些描述是由 Bl ackl er 和 Br own 在 1975年提出的。研究背景 以前有几点关于工作满意度到底意味着什么的看法。长期以来，有兴趣的调查

42、员就一直在研究措施使员工对工作更满意。从 1951 年起，Ka hn & Mor s e 就开 始研究影响工作满意度的外在因素与内在因素，并于 1973 年在美国工作时创办 了一个论坛。那次辩论的立场是呼吁利用越来越多的新措施来具体针对各项工作 的特点，如投影措施。我们应该使用假设性的案件或者以其他方式或间接的问题 接近受访者，因为直接的问题可能过于威胁性而使得受访者变得更加谨慎。这次 辩论虽然激烈，尽管有如调查工作条件这样严谨和仔细的研究，但在很大程度上 问题仍然悬而未决。从密歇根大学的社会研究所（I SR）在 1970 年对工作满意度的规模单一添加 剂的办法的比较中令人惊讶地发现：70 年

43、代的采用单一措施的整体工作满意度 的水平低( He r r i ck&Qui nn，1971) 。其他一些关于工作满意度的最近讨论，在完全避免任何的测量或定义的辩论 中，反对者更直接上诉潜在的社会问题。举例来说，部分支持工业的人文主义在 推定报告的基础上所产生的目前可用的工作满意度的绝对数字可以推断出美国 工人的不满（Ros ow，1974）。这种使用甜甜圈与孔的方法的不满，被界定为一 种重要的社会问题，因为工人工作满意度低的报告已经数以百万记。存在的问题 尽管有这些国家的最高权威，美国工人的工作满意度指数仍然只是一个很小的百分比。在盖洛普或者其他国家的调查中，这些指标是否由单一项目所采取的

44、措施或者是非常复杂的多个项目的规模，例如那些已经提到的对这一比例仍旧莫 名其妙的感到满意的工人报告。统计的事实是，要承担不论在何种程度上测量所 带来的复杂性。事实是不论是否研究使用数据的具体的工人在组装厂或国家随机 抽样，80%或更多的接受调查的美国人都满意他们的工作。I mbe r ma n 和 De f or es t 在芝加哥报道，一项由咨询公司对五家工厂的 3800 名员工的调查显示，79- 85%的员工对装配生产线的工作感到满意（I mbe r ma n， 1972）。Rut ger s 大学医学院的研究人员报告称，在 1968 年调查的 576 名联合汽车工会成员中，95的工人对他

45、们在汽车工厂的工作感到满意（Si a s s i ，Cr ocet t iSpi r o，1974 年）。虽然这些调查者称，他们的调查样本是工作满意度和工作 生活质量：对一个在巴尔的摩预付联盟健康计划的联合汽车保险集团代表的重 估，但是，样本应该被定格为约 13 年的服务期，平均 40 岁以下的，并每年赚取 九千美元以上的白人男子。这些结果一直以来和全国性的更仔细取样的调查报告是一致的。举例来说， 一项于 1954 年由芝加哥科学研究联营公司对 50 万受访者的全国调查称， 81的受访者对他们的工作感到满意。最近盖洛普组织报告称，在 1964 年的调查者 中 87的工人对他们的工作感到满意，1

46、973 年调查中的 77的工人对他们的工 作感到满意。一项由 I SR 调查中心最近做的调查的结果，显示 91的男性工人 对他们的工作感到满意。这些最近的调查结果和早前的调查在一个水平线上（这 是总结了 Ka pl an1958 年至 1973 年间的研究成果，1976 年）。这些不同的研究都明确地表明，多数的美国工人对他们的工作感到满意。这 些结果还显示了，在 1954 年至 1973 年间的一些小改变。这展现了工作满意度从 低到高约 79到约 95的一个转变进度。这种趋势显然不完全是 Che r ns（1975） 在联合国于最近的工作满意度研究讨论中指出的美国现象。这一模式得到了在各种不同

47、的组织中系统收集到的约 20 万名各个阶层（非 主管管理）的员工数据测试的额外支持。美国密西根大学科学知识（CRUSK）开 端研究中心 1966 年和 1970 年间在 15 个美国公司中约 33 个工厂和办事处收集到 这些数据（CRUSK，1970 年；Ta y l or & Bowe r s ，1972 年）。这些组织在各个维 度方面如：管理、经济状况，以及规模、衣领的颜色、技术等都是大不同的，虽 然他们没有被系统地采样。整体而言，报道称 20 万人中的 85的工人对他们的 工作感到满意，更多的人不感到不满意，但这些组织的统计范围是相当狭窄的。 一个 200 人的保险公司有 95%的员工对

48、他们的工作感到不满，该公司在 33 个公 司中排名首位。440 人的造纸厂有 76的工人对他们的工作感到不满意，这一公 司排在 33 个组织中不满意度最低位。尽管不是从美国组织中选取的科学样本， 满意度调查的范围也不是不同于全国性的调查。同时，我们发现在这些所有不同组织中这个大比例的员工，表示对他们的工 作没有感到不满意。像我们预测一样，我们发现工作满意度的可靠措施，并不是8和这些组织的旷工高度相关的。与怨气和营业额的措施（如少有数据可查的），人际关系与工作满意度并不是一贯的高或低。如果我们看看之间的分歧，工作组 在这些组织之间的因果关系的满意度和组织行为都不甚高（Ta y l or & Bowe r s ， 1972 年，p7779，89）。这些研究结果也许可以解释，无论我们在此基础上如 何界定以及对工作满意度多么仔细分类包括单独和重组，我们仍然只是对工作本 身进行衡量。我们不能期望来衡量所有要工人摆脱其内部的抗辩或期望的