Implementation

ity and to maintain the basic principles of the co-citation method, we utiliz

s. One is derived from the Acedo, Barroso and Galan’s meth

r, whic

3.2.1. Paper Collection

For reasons of objectiv

e an ad hoc heuristic method to find out the total papers that are topic related and then collect as many core papers as possible.

We use two ad hoc heuristic method

od (2006). It is a ‘snowball’ process that we extend our dataset of core papers from a small core. The other method is to search in Google Scholar (http://scholar.google.com) for related papers by using keywords like ‘strategic alignment’ and ‘alignment’. Finally, we use the number of citations listed by Google Scholar for the papers collected by the above methods. All of the data (number of citations per paper) was downloaded from Google Scholar on March14, 2007.

In the first method, we use three initial core papers: (1) Miles and Snow’s (1978) pape h defines three strategic types of organization, Defenders, Analyzers, and Prospectors, to explain how organizations move through the adaptive cycle and the methods they adopt to solve their entrepreneurial, engineering and administrative problems. (2) Sabherwal and Chan’s (2001) work, which examines the impact of alignment on perceived business performance using Miles and Snow’s popular classification of Defender, Analyzer, and Prospector business strategies. (3) Henderson and Venkatraman’s (1993) study, which argues IS strategic planning should be a concurrent activity that allows the potential of emerging technology to directly influence the strategic direction of the firm. Henderson and Venkatraman developed the Strategic Alignment Model (SAM) to identify two integrated types between business and IT domains. The three papers are important to the theoretical development of business and IS strategic alignment and they all have large number of citations for helping the analysis. Prior to March 14, 2007, they had

been cited 1046, 75, and 446 times, respectively.

From the reference papers of three initial core papers we use and papers citing them, we can gathe

p alternately until we have enou

r further possible core papers. If these new papers comply with the rules for selecting core papers, they are added to the core paper dataset and used to find out other possible core papers from the references of these new core papers and papers citing them.

We run the paper collection step and core paper selection ste gh core papers.

Figure 5 Paper Collection Processes Sabherwal and

Chan (2001) New core paper

New core paper

Henderson and Venkatraman (1993)

New core paper New core paper

New core paper

Miles and Snow (1978)

New core paper

If this paper is selected, we will check the reference of it and papers cited it.

Reference papers Papers citing this one

3.2.2. Core Paper Selection

papers are:

To avoid having too many papers, especially papers

2. . Obviously,

3. n analyze

Aft s 1 and 2, we had 44 core papers, as shown in Table 5.

The rules for selecting core

1. Papers from SCI or SSCI journals:

that are not good enough, the first rule is that a paper must have been published in an SCI or SSCI journal to be considered as core paper. If the article is from a book, we use criteria two and three to decide whether to include it the core paper pool.

Times cited: The number of times a paper has been cited represents its quality

newer papers will have had fewer chances of being cited than older ones. Therefore, the number of citations required for a paper to be considered a core paper depends on when it was published. We divide publications into four periods: the 1970s, 80s, 90s, and since 2000; and the minimum number of citations is 50, 40, 30, and 20 respectively.

Topic relevance: The only subjective criterion is the topic’s relevance. If we ca

total papers in the area of business and IS strategic alignment, the problem of subjectivity will be resolved. In other words, the more topic-related papers we can analyze, the less subjective the results will be. To apply this criterion, we have to check papers manually. To do this, we check the subject, abstract, and even the content of each candidate paper.

er completing Step

Table 5 Core paper list

Avison D, Jones J., Powell P. Wilson D. 2004 Journal of Strategic Information Systems 13(3): 223-246 Boynton AC, Zmud RW 1987 MIS Quarterly 11(1): 59-71

Broadbent M, Weill P 1993 IBM Systems Journal 32(1): 162-179 Broadbent M, Weill P, Clair DS 1999 MIS Quarterly 23(2): 159-182 Brown CV, Magill SL 1994 MIS Quarterly 18(4): 371

Burn JM, Szeto C 2000 Information & Management 37(4): 197-216 Chakravarthy BS 1987 Strategic Management Journal 8(6): 517-534

Chan YE 2002 MIS Quarterly Executive 1(21): 76 - 112

Chan YE, Huff SL,

Barclay DW, Copeland DG 1997 Information Systems Research 8(2): 125-150 Chester AN 1994 Research Technology Management 37(1): 25-32 Croteau AM, Bergeron F 2001 Journal of Strategic Information Systems 37(1): 25-32

Earl MJ 1989 NJ: Prentice-Hall Inc

Grant RM 2003 Strategic Management Journal 24(6): 491-517 Henderson JC 1990 Sloan Management Review 31(3): 7-18 Henderson JC, Sifonis JG 1988 MIS Quarterly 12(2): 187-200

Henderson JC, Venkatraman N 1991 Business Quarterly 55(3): 72-78 Henderson JC, Venkatraman N 1992 New York: Oxford University Press Henderson JC, Venkatraman N 1993 IBM Systems Journal 32(1): 4-16 Jackson PJ 1999 Information Systems Journal 9(4): 313 Johnston HR, Carrico SR 1988 MIS Quarterly 12(1): 37-48

Johnston HR, Vitale MR 1988 MIS Quarterly 12(2): 153-165

Jordon E., Tricker B. 1995 Journal of Strategic Information Systems 4(4): 357-382 Karimi J, Gupta YP, Somers TM 1996 Journal of Management Information Systems 12(4): 55-88 Kearns GS, Lederer AL 2000 Journal of Strategic Information Systems 9(4): 265-293

King WR 1978 MIS Quarterly 2(1): 27-37

Konsynski BR 1993 IBM Systems Journal 32(1): 111-142 Luftman JN, Lewis PR, Oldach SH 1993 IBM Systems Journal 32(1): 198-221

Luo Y, Park SH 2001 Strategic Management Journal 22(2): 141-155 McLean ER, Soden JV 1977 McKinsey and Company

Miles RE, Snow CC 1978 New York: McGraw-Hill

Palmer JW, Markus ML 2000 Information Systems Research 11(3): 241-259

Pyburn PJ 1983 MIS Quarterly 7(2): 1-14

Reich BH, Benbasat I 1996 MIS Quarterly 20(1): 55-81 Reich BH, Benbasat I 2000 MIS Quarterly 24(1): 81-113

Sabherwal R, Chan YE 2001 Information Systems Research 12(1): 11–33 Sabherwal R, Hirschheim R, Goles T 2001 Organization Science 12(2): 179-197 Santhanam R, Hartono E 2003 MIS Quarterly 27(1): 125-153

Segars AH, Grover V 1999 Information Systems Research 10(3): 199-232 Teo TSH, Ang JSK 1999 International Journal of Information Management

19(2): 173-185

Teo TSH, King WR 1997 Journal of Management Information Systems 14(1): 185-214 van der Zee JTM, de Jong B 1999 Journal of Management Information Systems 16(2): 137-156 Venkatraman N 1989 Management Science 35(8): 942-962

Venkatraman N, Henderson JC,

Oldach S 1993 European Management Journal 11(2): 139

Wiseman C 1988 McGraw-Hill Professional

3.2.3. Producing a Matrix

Producing a 44 x 44 matrix is time consuming and prone to error because it is necessary to check every two papers’ co-citation paper and count it. For example, we check the papers that both cite Chan (2002) and Chester (1994), and the number of these papers is the co-citation count of Chan (2002) and Chester (1994). To overcome these two problems, we build a database to handle process of producing this matrix.

First of all, we arrange the 44 papers and assign them numbers, after which we build two database tables. The first is a “citation” table that contains three fields: 1) “CorePaperSerial”, which details the serial numbers of 44 core papers; it is the primary key of this table; 2)

“SubPaperSerial”, which lists the serial numbers of papers that cite one of the core papers. 3)

“SubPaperSubSerial” are redundant values of “SubPaperSerial” that will be updated after the next step. The second table is called the “subpapers” table. It stores the data about papers in the citation table, such as author, topic, and year of publication.

The next step is to sort the data in the “subpapers” table using the topic as the sort criterion.

After sorting, we compare the author, published year, and topic of each paper with every other papers. If they are the same, the value of “SubPaperSubSerial” field of the “citation” table will be updated to the same number. According to the value of “SubPaperSerial”, we know the paper of this “SubPaperSerial” cite the paper of “CorePaperSerial”. From the same value of

“SubPaperSubSerial”, we have many “SubPaperSerial” values and know that these papers are the same one. Moreover, we can find out papers of “CorePaperSerial” are cited by these papers of

“SubPaperSerial” and actually the papers of “CorePaperSerial” are cited by the same paper.

Finally, we count every two core papers’ co-citation number by comparing the

“SubPaperSubSerial” field.

By using this method, we can avoid using complex SQL commands or spending a lot of time on computations..

Citation

CorePaperSerial SubPaperSerial SubPaperSubSerial

Subpapers Serial Author Topic

PublishedYear

Figure 6 Database Schema

3.2.4. Transformation to a Correlation Matrix

We collect the co-citation counts (when two documents are cited by the same paper, then count plus 1) and transform the co-citation matrix into a Pearson correlation matrix.

There are two advantages in transforming the raw co-citation matrix into a correlation matrix: (i) the data is standardized, and (ii) the number of zeros in the matrix is reduced. The correlation matrix shows that the relationships of the following five papers are weak with the other papers. Therefore, we drop them.

Avison D, Jones J., Powell P. Wilson D. 2004 Journal of Strategic Information Systems 13(3): 223-246 Chester AN 1994 Research Technology Management 37(1): 25-32 Jackson PJ 1999 Information Systems Journal 9(4): 313

Jordon E., Tricker B. 1995 Journal of Strategic Information Systems 4(4): 357-382 Luo Y, Park SH 2001 Strategic Management Journal 22(2): 141-155

We now have 39 core papers for statistical analysis