Demand for More MIS Empirical Research

全文

(1)Demand for more MIS Empirical Research. Demand for More MIS. F~mpirical. 49. Research. Heng-Li Yang (m-~~O. Department of Management Information -Systems. College of Business Administration. National Cheng-Chi University. Taipei, Taiwan. Abstract This paper reviews the current status of MIS research and describes the research methodologies. The empirical research is emphasized. The sta tus of empirical research in Taiwan is discu&')ed, especially from the gradu ate program training. It concludes with the suggestion that more emphasis should be placed on empirical resarch in Taiwan. ......;~i:. Keywords: Empirical Research; Research Methods. 1.. Introduction. methods. Section 3 emphasizes the importance of empirical research. Section 4 di&---usses the current. The Management Information Systems. situation of empirical research in Taiwan, especial. (MIS) discipline is referred to as Information. ly from the &')pect of graduate education. Section 5. Management (1M) in Taiwan. 1t is about. concludes this paper with the suggestion.. fourteen years since the first 1M undergraduate department was established at Fu-Jen Univer. t. t The Immature MIS Held. in 1981. However, in local conferences or. According to Keen's definition (25) ,. journal, there are few papers discussing the. MIS is "the effective design, delivery and use of information systems organizations". MIS research is "the systematic investigation of the. MIS research and its methodologies. The pur pose of this paper is to provide such a discussion.. development, operation, use and/or impact of The remaining of this section briefly reviews. an information (sub)system in an organization. the Current status of MIS field. Section 2 provides. al environment" (21) . A field is a common. the MIS research cycle and describes the method. ground on which members define the area and. ologies, especially introduces empirical research. its important measures (9) . An reference dis. •. II I. I. II. .1.1._.

(2) ...... 50. ~1i~'. cipline is an established field to which one. in which deduction is applied to test theories so. looks to get an idea of what good MIS research. as to create new experiences or observations. It. would look like, if one could ever do it (25J .. is a theory-testing and mainly relies upon quan. MIS has many reference disciplines: computer. titative, empirical methods.. science, management science, management, organization behavior, political science, eco. In the literature, there are a number of. nomics, accounting, etc. Mis field has been. papers discussing MIS re;;ea;ch methodologies,. argued as a "fragmented adhocracy" (5) , criti. e.g.,(1,2,7,8,9,12,19,20,21,23,24,28,29,. cized as no cumulative tradition and no theoret. 30,31,32,33,34,36) The MIS research can be. ical base (25) . Even up to now, Cheon et al.. classified to three categories in.Figure 2:. (9) assess the changing maturity of the MIS. 1. Modelling . A model is an abstraction of re. field over the past 10 years and conclude that. ality. Modelling is the process of creating. there is little indication of change in-maturity.. the abstraction that preserves the essential. 2.. !\'lIS Research Methodologies According to Kolb et al. 's "learning cycle". (26) , learning might start with the experience. properties of a complex system, but sup presses much speific details. A theoretical model includes (1) a set of constructs;. laws of interaction. of. the. constructs;. of an event or stimulus - . - concrete experi-'. boundaries within which the theory is ex. ences, which the individual then observes it. pected to hold; (4) system states; (5) propo. and reflects upon in trying to make sense of it.. sitions and hypothesis (15) .. This might lead to formation of abstract con. 2. Implementation.. In implementation, a. cepts and generalization, which testing in new. prototype system is designed, constructed. situations creates new experiences that enable. and used on the basis of the theoretical model.. consequent reflection, observation, and ulti. 3. Empirical Studies. In empirical studies,. mately new rules. Thi<; learning cycle describes. the unobservable, abstract concepts are op. how hmnan beings learn and is similar to Straub's. erationalized into observable empirical mea. "scientific research cycle" (31) though the dia. sures; the data are gathered and analyzed in. grammatic representations are different.. an attempt to verify the hypothesis. The purpose is to check the impact of the above. In Figure 1, the right hand side of the cy cle corresponds to exploratory research, in which induction is applied to construct expla. constructed information system on a person, group, organization, and society. The methods (such &<; thinking) of relating to. nations and theories. It is a theory-building. modelling and implementation are hard to be de. process and mainly relies upon qualitative,. scribed precisely and taught. However, there are. non-empirical methods. The left hand side of. some rigorous require-ments for conducting empir. the cycle corresponds to confirmatory research,. ical studies.. .A.

(3) Demand for more MIS Empirical Research. Concrete Experiences. /~. Testing Implications of Concepts in New Situations (Confirmatory Research). Observations and Reflections (Exploratory Research). ~. /. Formation of Abstract Concepts,. and Generalization. (Conceptual Refinements). Figure 1: The Learning Cycle. the basis for Scientific Reseach. Empirical Research. Modelling. ~/. Implementation. Figure 2: The MIS Research Cycle. Conceptual Level:. Unobservable, Abstract Concepts. (Conceptual Definitions for Constructs). Opera tionaliza tion Process Empirical Level:. Observable Indicators or Measures. L I - - ______________________________________________ (Operationalization Definitions for Constructs) Figure 3: Operationalizing Constructs. I. ill.. ~. 51.

(4) 52 f<. 2. t. Empirical Research. ferences on the characteristic one is attempting. to measure and nothing else. A measure is re. An important stage in empirical studies is the. liable to the extent that independent but com. operationalization process; that is, translation. parable measures of the same construct of a. of abstract concepts into indicators or measures. given object agree (10J.. that enable observations to be made. How do we measure an abstract concept, such as, User. Assume that:. Information Satisfaction (22J? It is impor. Xo is the observed score;. tant to have both conceptual definitions and. X T is the truce score;. operationalization definitions for each construct. al level and to have validated instrument to. Xs is the systematic source..;; of error;. and X R is the random sources of error.. then XO=XT+ Xs+ X R. If X R = 0, the measure is reliable.. measure them, as shown in Figure 3.. If both X R =. in empirical studies. The two key points are to define the constructs precisely in the conceptu. °. and Xs. 0, the measure is. valid.. Instrument Development The instrument. So, reliability is a necessary condition. development is a sophisticated and iterative. suffcient condition fofor validity. _. not a. process. It starts with specifying domain of. constructs, generating sample of items, col. The reliability coefficient is mathematically. lecting data for pretesting, purifying measure,. defined as the ratio of true score variance to ob. collecting new data for pilot testing, assessing. served score variance (11 J • There are several. reliability, assessing validity, and developing. types of reliability coefficients, such as coeffi. norms (10J.. cient of consistency, coefficient of stability,. coefficient of equivalence,. etc.. (11,14J.. Reliability and Validity Reliability refers. There are also several types. to the degree to which observed scores are free. as, content validity, criterion-related validity,. from errors of measuremet. Validity refers. construct validity, internal and external validi. to the appropriateness, meaningfulness, and. ty (4, 11, 14J.. validity, such. usefulness of the specific inferences made from the measures (I 4 J. It refers to the usefulness. Empirical Research Methods There are a. of inferences drawn from test scores for a given. number of empirical research strategies or. purpose under a prescribed set of conditions. methods. Based on the methods classified by. (l1J. That is,. (1, 9, 20,30, 34J, we have the following.. • Laboratory Experiment. A laboratory ex. in general, refers to the ex. tent to which a claim or conclusion is based on. Iii. sound logic (13J. A measure is valid when the. periment is taken place in a unnatural setting. differences in observed scores reflect true dif-. laboratory, in which the experimen trer.

(5) .......... Demand for more MIS Empirical Research. has control over the assignment to groups, controls and manipulates independent vari. 53. measured); action(participative) research (the researcher is a participant in the imple mentation of a system); longitudinal study. and measures dependent variables.. (behaviour is measured at a number of prints in. • Field Experiment. A field experiment. IS. time during a finite period of tmie); and. place in a natural setting, in which the. archival (document) research (primary or. eXJ)erlm.enter manipulates independent vari. secondary documents are examined), etc.. while exerting a<:; much control as the. The point is that no strategy is more ap. situation permits over other, possibly con. propriate than others for all research purposes.. founding variables, and measures dependent. For. case studies are useful in formu hypotheses, not testing hypotheses; lab. variables.. oratory experiments are usful in maximizing in. • Field Study. A field study is ex post facto in nature. It is taken in a natural setting, in no maepenaent vanalJles are manipu lated while measuring dependent variables.. validity. not external valicH.ty. The se lection of research strategies should be guided by purpose of the research, resources available to the researcher, and other considerations. (30) .. • Sample Survey. A sample survey is taken. There are some data collection methods,. from a large sample in a natural setting. No. such as interviews, questionnaires, and obser. manipulation can be done though influences. vation. which might be used concurrently in a. of confounding variables are "controlled". selected. statistically.. 3. • Case Study. A case study examines a phe nomenon in its natural. employing. multiple methods of data collection. to. gather. information from one or a few entities (peo groups, or organizations) (3). The re. strategy.. The Importanre of EmpirirnJ Research usage of. methods is one of evalua. tion criteria by which Cheon et al. (9) apply to assess the maturity of MIS field. 1 They claim that if MIS field could become more mature, research methods that. phenomena through. searcher may have little or no priori knowl. hypothesis testing would. applied more often. edge of what the variables of interest will be. than research methods that describe or con. ([nd how they will be measured.. struct phenomena. All the methods related to modelling and implement-ation are for describ. There are other methods (4, 19. 24) •. ing phenomena - - exploratory research. Be. such as, simulation (a natural system and re. sides, there are some empirical research meth. al-world-like events are replicated and· depen. ods, e.g., case study, for exploratory research.. variables - - participants' behaviors are. ............

(6) 54. Modelling and implementation research. by Cheon et a1. (9), empirical research has in. certainly have their value, especially for basic. creased in US, Canada and Europe. Hamilton. understanding of a phenomenon, creating a. and Ives (20) reported that MIS research pub. model to capture the real-world system proper. lished in journal articles during 1970-79 em. ties and providing a prototype to show its feasi. ployed a surprisingly high (70.1 %) percentage. bili ty. However, there are too many "frame. nonempirical research strategies. However,. works" and "models", and too few "well-an. the percentage declined po';n to 43 % during. chored theories" in the MIS field (35). Initial exploratory research should be followed by the. 1977-85 according to the survey by Farhoomand (16) . In addition, 17.6% of the empirical. ory-building so as to contribute to the building. studies during 1970-79 adopted the laboratory. of a cumulative research tradition. Immple. experiment strategy (20); the ratio increased. mentation, the design and construction of a. to 23.3% during 1980-89 (9).. system, might be useful and is also an "intrin sically satisfying experience" (35). However,. 4. Empirical Research in Taiwan. Weber (35) wonder how much such work con. In Taiwan, the local MIS journals are in their. tribute to progress in the MIS field. He calls. infancy. The Information Management pub. as "the lure of design and construction". lished by the Society of Information Manage. snd suggests that the implementation should be. ment just had its first issue. Though there was. based on some theory of information systems. a high percentage of empirical re..<;earch papers. that can be used to predict hte likely success of. in the first issue (five out of seven), we cannot. failure of a design, and these predictions can be. conclude that empirical research gains its de. tested empirically.. served attention because the sample is still to few.. MIS research is the systematic investiga tion of the development, operation, use and!or. There are six graduate schools of information. impact of an information (sub)system in an or. management (1M) in Taiwan. National Sun Yat. ganizational enviromnent. Keen (25) suggests. Sen University (NSYSU, established in 1989),. "research-with-practice" - - the research should. Nationl Chiao-Tong University (NCTU, 1989) ,. point towards practice. We need to apply em. Nationl Central University (NCU, 1991), Nation!. pirical research methods to understand what. Cherg-chi University (NCCU, 1992) , Tam-Kang. real impact of the implemented system or pro. University (1992), and Da-Yeh Institute of Tech. posed model on an individual, group, organiza. nology (DYIT, 1993) though some graduate. tion, inter-organizational environment, or so. schools of management or busine&<; administration. ciety.. offer MIS major in their master programs. Three. Though there is still very little change in. of universities (NSYSU, NCTU and NCU) have. maturity with respect to the criteria that used. begun Ph.D. programs since Fall 1994. Others. r.

(7) Demand for more MIS Empirical Research. only provide ma<;ter program<;.. 55. Idally, it would be a two-semester course. The. In the following, the empirical research in. first semester covers the measurement theory,. Taiwan will first be observed from two phe. research strategies and data collection methods. nomena in MIS graduate progaram. Then, pa. (and introduces some analysis teehniques if. pers published in annual Conference on Infor. time permits). The second semester could. mation Management, Taiwan will also be ex. cus on the training of research proposals and. amined. The data about Taiwan are from this. parper reviews. There would be some related. research and are based on the archives.. courses. If we consider this "empirical research course" as the middle, its upper would be some. The Phenomenon 1. the "Research. discussions on. research methods (no matter. Methods" Course In U. S. and Canada, MIS. what empirical or not) from several philosophi. graduate programs offer master degree which is. perspectives; its lower would be some ad. either MBA-oriented or MS-oriented (thesis. vanced statistics or analysis courses.. required). A MS-oriented student is usually re quired to take the "Research Methods" course. The. course. contents. of. the. "Research. Methods" should include the following:. Graduate education in Taiwan is more MS-oriented. However, currently only NCU, Tam-Kang University and DYIT have required their graduate students to take "Research. • To introduce the student the following. Measurement Reliability. Methods" course (one semester). Others offer it as an elective coursel. •. Measurement Validity. Different Data Colleetion Methods. ~. Different Research Strategies. The Phenomenon 2 - - the Master The ses The first graduates in NCTU and NSYSU were 1991, in NCU were 1993, and in NCCU. • To give the student a preliminary training to. and Tam-Kang University were 1994. DYIT. write a research proposal and give criticisms on. has not had graduates yet. During the period. papers.. between 1991 and 1994, there were 258 master theses. This research has reviewed their ab. It is not a course in data analysis or statis. stracts and the seetions of resarch methods.. so no particular expertise in statistics will. Also, the particular sections or chapters de. be llSsumed, although familiarity with some. scribing empirical studies (if any) were quickly. statistical analysis techniques would be helpful.. browed through.. 1 NeeD first offered this course as an elective course in 1993. did not provide it in 1994. but will switch to treat as a requisite course beginning in Fall 1995.. ............ ''''110,.

(8) ~. MliM. -ftftli~+=~. A", shown in Table 1, only 66 (about. (1); others applied the sa;ond choice [9, 16,. 25.58 %) theses were related to empirical re. 20, 21) . Therefore, the second choice are also. search. Others (about 74.42%) were related. adopted here for comparison.. 2. to modelling and implementation Table 2 further classifies I. :. A brief description of the Ives et al.. •. 66 empiri. framework of IS research is described. cal master theses by research methods (strate. Readers should check their paper (21) for de. gies) and types.. tails. It includes three groups ''bf variables:. I. I'. The classification of. I. I. IS. 1. Environmental Variables include five. based on Section 2. 1.. classes of environmental varial,>les:. in literature [20, 21, 34) do not distinguish. organizational, user, IS development, and. ~'sample. IS operations.. surveys" from "field studies". For two cate. 2. Information System Variables include. gories. However, we should note that the key. three classes of IS variables: IS content,. difference between sample surveys and feild. presentation form, and time of representa. studies is that in the typical field study the. tion.. comparison the table combines. "sample" is not a random one, and the popula tion to which generalizations might be made is. 3.Process Variables includes. classes of. variables: development, operation and use.. left un~pecified; while in the typical sample survey influences of compounding variables are. Based upon thexe three variable groups, five. controlled" statistically (30) . According to. categories of IS research topics are derived:. this strict distinction. all of those theses in column. 1. Type I: involves variables within a. 4 of Table 2 are. studies, not sample surveys.. variable group from. of three. groups: environmental. process, or IS. About research types,. are two choic. For example: testing hypothesis that data. es in the literature: one is based on the key. entered online (IS variable) will be more. word classification. of Barki, Rivard,. accurate (IS variable) than data entered of. and Talbot [ 6); the second is based on the. . or testing hypothesis that IS delivered. framework of information system (IS) research. within budget (development process variable). framework presented by Ives, Hamilton, and. will be perceived by user as being of a higher. Davis (21). For those reviews of MIS research. quality (use process variable) than those. in the literature, only one used the first choice. are above budget.. 2 Among these 192 theses,six had short "empirical studies" sections at end of their theses. However, they just used very few artificial data to try their models. This paper does not count them as "emDirical research".. ,~.

(9) Demand for more MIS Empirical Research. lVleXlelllTIlg & . 57. Empirical. Studies. 58% Table 1: MIS Master Research in Taiwan During 1991-94. l. Research Methods Laboratory Experiment. Recearch Types. T~pe. 1~ i. Th~. Percent. Experiment. I: A Single Variable 2. Type II: Relationship I between Environment & 1 Process Variable Groups I. .129. 43.94%. T~'pe. III: Relationship between Process & IS o Variable '-' 1 'UU~R!> T~'1)e IV: Relationship between Environment & 1 IS Variable Groups Type V: Relationships among all Groups of Variables. o. o. 2. o. 1. 1. o. 3. 4.. o. 5. 7.57%. 3.03%. I. 11. o. 3. 11. 16.67%. Total Percent. \lote that among these 66 master theses: having explicit hypotheses: 25 having pretest: 5 having discussed reliability and validity: 14 having operationalization definitions for variables: 18. \lote that among 13 master theses applying case studies, 5 were single-case studies. Table 2: Empirical MIS Master Research in Taiwan During 1991-94.

(10) 58. I. :il:gwp~ffil. ~liM. -fLfLliif+=Jj. I. II I. 2. Type II: examines the influence of one or. method. No one were field experiment. Com. more variables from the environmental vari. paring with the review during the period 1980. ables group on the process group measures.. to 1989 reported by Cheon (9) (among 463. For example: testing hypothesis that organi. empirical articles they reviewed, 59.6 % were. zations located in rural areas (external envi. field studies, 16.2 % were case studies, 23.3 %. ronmental variable) will incur signifcantly. were laboratory experiments, only 0.9% were. higher system operation cost (operation process. field experiments) , the distribution pattern. variable) than will their urban counterparts.. were somewhat similar. However, our labora. i. II I:. I I. ~. I. '. II. I. 3. Type III: focuses on the influence of IS ,. .. tory experiments were still too few, and the. I. variables on process variable measures.. I. II. For example: testing hypothesis that an on line IS (IS variable) will require more time. II. II i i. II. field experiments were lacking. Among the 13 case studies, 5 (38.46 % ). to develop (development process variable). were single-case designs. As discussed by Ben. than an offline IS.. basat et al. (3), multiple-case designs are de. t Type IV: examines the rei a tionshi p be. sirable when the intent of the research is de. tween environmental variables and IS vari. scription, theory building, or theory testing.. ables.. Multiple-case designs allow for cross-case anal. For example: testing hypothesis that organi. ysis and the extension of theory.. I. zations facing higher uncertainty in the mar ketplace (external environmental variable). The questionnaire~ as data collection. will have a shorter reporting interval (IS. method were the most common in those field. variable) .. studies or laboratory experiments. However,. 5. Type V: examines relationships between. few took pre-tests in Taiwan. Only some gave. variables from each of the three groups.. operationalization definitions for construct vari. For example: testing hypothesis that there. ables. The reliability and validity issues of. will be an interaction effect between the IS. questionnaires were seldom addressed. That is,. I'. presentation model (IS variable) and user psy. most of those instruments were not validated.. II. chological type (user environmental variable). I. II. on decision quality (use process variable).. In addition, among those 66 theses, only. I. 25 gave explicit hypotheses to test. Without. III. Research Methods As shown in Table 2, among these 66 theses, 37 (56.06 %) of. get lost. Unless their works were in explorato. them were field studies, 13 (19. 7 %) were case. ry or hypotheses generation stages, the contri. studies, 11 (16.67%) applied archival research. butions would be unclear.. hypotheses, the themes of theses would easily. (use secondary data to test their models), and on ly 5 (7.57%). took laboratory experiment. Research Types As shown in Table 2,.

(11) ..... Demand for more MIS Empirical Research. re~earch ~tudies,. orous requirements for conducting empirical. types I and II had most other types (III, IV and v ) were. atively. popular. The distribution pattern. were similar to the review during the period to 1989 reported by Cheon (9J (among articles they reviewed, the per centage.') of. 59. I, II, III, IV, and V were. studies. Without appropriate training, 1M students would not know how to proceed empirical studies. In addition, there is a seri ous problem of lacking appropriate high percentage of. A ( even. some graduating from MIS. 52.1 %, 25.5%, 14.7%, 2.8%, and 5.2% ,. might not know what contents the "research. re~pectively) .. method" course should contain.. Discussions for Few Empirical Stud. • Empirical research might need much resoures.. ies There are several possible reasons for this. To be a "quality research", an empirical. thesis research phenomenon.. study needs a sophisticated and iterative pro. • 1M in Taiwan is more computer science. cess to develop its instrument. In addition,. oriented. As observed. no matter what kind of research methods are Farn (17J in 1992, the fac. applied, the requirements of time and mone. protiles for most of 1M departments in. tary resources on researchers cannot be avoided.. Taiwan are quite imbalanced: most have. In comparison, modelling or implementation. very strong computer science, industrial en. in the Taiwanese ma')ter level might not re-. gineering, and operation research mix, and lacking management background in general and MIS background in particular. Liang. (27J also pointed out: graduate students in Taiwan are more. in "infor. personal resources or might ap resources provided by universities.. • The linkage between academy and in dustry is weak. An empirical research needs respondent's (or. mation technology", but have much less. interviewee's) cooperations. It would be. sense in "busine&<;" or "management". This. even better if the copperations come from in. are owing to historical back and in general have little changes even to date. As a result, some facuty mem still only conduct purely computer sci ence type of projects.. dustry - - field studies or even field experi ments. However. in Taiwan, the links be tween 1M academics and practitioners are weak. Examples of business sponsoring re search proiects are. • No appropriate research method training. research methods course has not been. Papers in Annual Conference on Informa tion Management, Taiwan To get some. given much attention in Taiwan. As also. crosschecking, the papers in annual Confer. mentioned in Section 2, there are some rig. ence on Information Manegement, Taiwan 8. As mentioned in the above phenomenon 1, t he. I ~.

(12) 60. jl~wpgifH. m1iW3. -:tL:tL1i~+=Fl. II ,. I. 'I. are also examined here. However, we should. what and how the impact of information sys. note that (i) the different nature of journals. tems on organizations and society in Taiwan.. and conferences (the paper acceptance rates and in. The 1M graduate program in Taiwan should. formality are much higher in conferences); ( ii ). strengthen the "Research Methods" course. one issue of the conference proceedings only. training to teach students how to conduct em. had abstracts, not full papers, i. e., those 57. pirical research and how to review other's pa. I. I. I I. I. papers in the second conference (1991); (iii). pers on empirical studies. fAcademics in 1M de. some papers in conferences were in fact the re. partments should recognize that they belong to. sults of previous master theses.. business colleges, and research-with-practice is. I. I. I. As shown in Table 3, only 35 (about. key to establish the 1M discipline features and. 19.77 %) theses were related to emptrical re. to contribute MIS research. More research. search. This number is slightly lower than the. should be on the effectiveness, utility , and. ratio (25.58 %) in master theses during the. management aspects of information systems.. same period. Others (about 80.23 %) were re. There might be a need for faculty retraining. lated to modelling and implementation.. and development programs. Alternatively, 1M. I. Ii. I. I', I. ill. I. ! I. I. '!. I. .1. .. Similar to Table 2, Tables 4 also further. departments might recruit more faculty mem. II. classifies these 35 empirical papers by research. II. methods and types. Comparing these two ta. ,I. bles, we can find that the distribution patterns. It is essential to consider the reliability. of the papers and master theses among research. and validity of an instrument. The instrument. methods and types are quite similar.. should be pre-tested, pilot tested, and validity.. i. I. 5. i. I. i,l I. Conclusions and Suggestions. bers with management or MIS degree.. Without validated instruments, the results of empirical studies would not be convincing.. MIS field is still immature. Greater emphasis. Case studies are valuable, but a single case used. should be placed on testing theories and con. for exploration may be better followed by a. structing empirically based theories.. multiple-case study. Graduate students should be instructed to write down their hypotheses. On the academical side, in order to con. and figure out how to operationalize the con. tribute more to this field, we need more empir. structs in empirical studies. Without these,. ical research in Taiwan. On the practical side,. the empirical studies would not be rigorous.. we also need empirical research to understand I. I. I. II II. 3 The conference were called National Conference on Information Management during 1990 (lst) to 1992 (4th) ,and called International Conference on Information Management sence 1993 (5th).. Ii Ii, !. II I. I. ~. Iii. ,. I. I. F. ,... I.

(13) - Demand for more MIS Empirical Research. I Total. Empirical Studie...,. 142 ) 80.23%. 1177 )100%. 135 I 19.77%. !. 61. 3: MIS Papers Published in Gmference on Information Management. Taiwan During. 1991-94. 1<esearch ~ethods Recearch Types. T _1. 'Y. I Field 'r._. E'. T~IM'. : Group. Single. '1\I)e II: Relationship between Environment & Proces..c;; Variable Groups. 4 0. 0. Field Study & Sample Survey. I,Study Case. 12. 2. 9. 2. I. Archival 1<esearch (Secondary Data). Total. Percent. %. 5 11. 31.43% --------. T~-I)t~. III: Relationship hetween Process & IS. 0. Variable Groups '1\I)e IV: Relationship between Environment & IS Variable Groups. 0. 0. 0. 0. 0. 0. 2. 0,%. 2. 10. 1. 2.86%. I. 0. 0 ------------. TH)e V: Relationships among all Groups Variables Total Percent ---------. 0. 0. 1. 4 11.43%. 0 0%. 22 4 5 62.86% 11.43% 14.28%. 0. 0. 35 100%. that among these 35 papers:. 6. having their abstracts, not. papers in Conference. 8 also appeared in Table 2, i. e., were master theses 11 having explicit hypotheses. 3 having pretest 6 having disc used reliability or (and) validity ;) having operationalization definitions Tahlp 4: Empirical MIS Papers Published in G_mlerence on Information Management, Taiwan During 1991-94. ~. ..

(14) ............ =.. - tLtL1i-4+ A. 62 . I t is recommended to try field experiments in Taiwan to test MIS theories in business or. '.. North-Holland, Amsterdam,. 1985, Pp.. 193-201.. ganizations. However, it would need more co. (8)Buckley, J.W. and Buckley, M.H., Re. operation from industries. It is also recom. search Methodology &3 Business Decisions,. mended to consider type I II, IV research, and. National A"lSOCiation of Accounts Publication,. finally the most sophisticated Type V research.. New York. 1976. (9) Cheon, M. J., Grover, V., and Sabherwal,. References (1) Alavi , M. and Carlson, P., "A Review of MIS Research and Disciplinary Development" , Journal of Management Information Systems ,. Vol. 8, No.4, Spring 1992, p. 45-62.. i. Ii. R., "The evolution of empirical research in IS. A study in IS maturity", Infor. mation &3 Management, Vol: 24, 1993.. pp. 107-119. (10) Churchill , G.A., "A Paradigm for Devel. (2)Antill, L., "Selection of a Research Method",. oping Better Measures of Marketing Con. in Research Methods in Informatiian Systems,. structs", Journal of Marketing Research,. edited by E. Mumford et al. , North-Holland,. Vol. XVI, February 1979, pp. 64-73. (11) Crocker, L., and Algina, J., Introduction. Amsterdam, 1985, pp.203-215. (3) Benbasat, 1.,. Goldstein,. D.K.,. and. Mead, M., "The Case Research Strategy in Studies of Information Systems", MIS. to Classical and Modern Test Theory,. HRW, New York, 1986. (12)Culnan,. M.. and. Seanson,. E.B.,. Quarterly, Vol. 11, No.3, September. "Research in Management Information. 1987, pp. 369-386.. Systems, 1980-1984: Points of Work and. (4) Bailey, K.D., Metlwds of Social Research, 3rd edition, The Free Press, Macmillan, New York, 1987.. 1986, pp. 289-302. (13) Dane, F.. (5) Banville, C. and Landry, M., "Can the Field of MIS be Disciplined?", Gnnmunications of the ACM, Vol. 32, No. 1, January 1989,. pp.48-60. (6)Barki, H., Rivard, S., and Talbot,. Reference", MIS Quarterly, September. Brooks/Cole. C.,. Research. Publishing. Metluxls,. Co.,. Pacific. Grove, 1990. (14) Dooley , D., Social Research Methods, 2nd edition, Prentice-Hall, Englewood. J., "An. Infonnation SysteIIL"l Keyword Oassification. Cliffs, 1984. (15) Dubin, R., "Theory Building in Applied. Scheme", MIS Quarterly, Vol. 34, No.. Areas", in Handbook of Industrial and. 4, June 19H8, pp. 299-322.. Organizational Psychology,. (7) Boland, R.J .• "Phenomenology: A Preferred. edited by. M.D. Dunnette, Rand McNally College. Approach to Research on Information Sys. Pub. Co., Chicago, Ill., pp. 17-39,. tems", in Research Metlwds in Infarmation. 1976.. Systems, edited by E. Mumford et al... (16) Farhoomand, A.F., "Scientific Progress.

(15) Demand for more MIS Empirical Research. of Management Information Systems" , Data Base, Summer 1987, pp. 48-56.. [17] Farn , C.-K.,. 63. June 1985, pp. 141-156. (24)Jenkins, M.A.. "Research Methodologies. "The Status of MIS. and MIS Research", in Research Methods in. Education in Taiwan" .' NACISPA Newsletter, October 1992, pp. 1-4.. Information S:yst.erns, edited by E. Mumford. Fitzgerald, G.. Hi rschheim, Mumford, E., Wood-Harper,. R.A. , A.T .•. et al., North-Holland, Am.<;terdam. 1985. pp. 103-117. (25JKeen.. P.,. "MIS Research:. Reference. .. Information Systems Research Methodolcgy:. Disciplines and a Cumulative Tradition" ,. An Introduction to the Debate" , in. Proceedings of the. Hrst. International. Research Methods in Information Sys. Conference on terns, edited by E. Mumford et al.,. North-Holland, Amsterdam, 1985. pp.. 3-9.. Information S:ystems, December 8--10, Philadelphia. PA, 1980. pp. 1-8.. (26JKolb, D.A., Rubin, I.M.• and McIntyre. a Paradigm. 1. M.. Organizational Psychology: An. for Information Systems Research", in. Experiential Approach. Prentice Hall.. Galliers, R.D., "In Search. Research Methods in Information Sys. L.ondon. 1979.. :;". terns. edited by E. Mumford et al .•. (27J Liang. T. -P., "A Comparison of IS. North-Holland. Amsterdam. 1985. pp.. 281-297.. Education Between The United States and. Hamilton. S.. and Ives. B... "MIS. Research Strategies". Information. {:3. Management, Vol. 5. 1982. pp. 339-347.. Ives. B., Hamilton, S.. and Davis.. Taiwan". NACISPA Newsletter. October 1992. pp. 4-7. (28J Orlikowski, S.l. and Baroudi, 1.1., "Studying Information Technology in Or. G. B.. "A Framework for Research in. ganizations: Research Approaches and As sumption..<;", InfOrmation S:ysterns Research,. Computer-Based Management Information. Vol. 2. No. 11, March 1991, pp. 3-14.. Systems", Managernent Science, Vol. 26. No.9, September 1980, pp. 910-934.. (29JPinsonneault, A.. and Kraemer.. K.,. Ives, B., Olson, M.H., and Baroudi, 1.. "Survey Research Methodology in Management Information Systems: An. J., "The Mea.<;urement of User Information. Assessment" , Journal of Management. Satisfaction", Communications of the. Information Systems. Vol. 10. No.2,. ACM, Vol. 26, No. 10, October 1983, pp. 785-793. [23J]arvenpaa, S., Dickson, G. W., and DeSanctis, G., "Methodological Issues in. Fall 1993, pp. 75-105.. (30JStone, E., Research Methods in Organi zational Behavior, Scott, Foresman and. Company, 1978.. Experimental IS Research: Experiences. (31JStraub, D.W., "Validating Instruments. and Recommendation", MIS Quarterly,. in MIS Research", MIS Quarterly, Vol..

(16) 1 ~. mnM-nnn~+=~. 13. No.2, June 1989. pp. 147-169. (32) Subramanian, A. and Nilakanta. S., "Measurement - - A Blueprint for Theory-Buiding in MIS", Information Management. Jan. 1994, pp. 13-20.. (35) Weber, R.,. "Toward a Theory of. Artifacts: A Paradigmatic Base for Information System Research" , Journal of Information Research. Spring 1987, pp.. (33)Teng. J. T.e. and Galletta, D.F., "MIS. 3-19. (36) Weick , K.E., "Theoreti,cal Assumptions. Research Directions: A Survey of Researchers' Views", Data Base. Fall. and Research Methodology Selection". in The Information Systems Research. 1990, pp. 1-10. (34) Van Horn, R.L., "Empirical Studies on Management Information Systems", Data. Challenge, Proceedings, HBS Research. (3. Base, Vol 5, No. 2-4, 1973. pp. 171. Collogium, edited by F. W. McFarlan.. Harvard Business School Press, Boston. 1984, pp. 110-132.. 180.. ~.

(17)