The large number of KMS functions were aggregated to the two measures exten- sion and intensity of the usage of KMS functions:
g(KMS) = group of [integrative; interactive; bridging] KMS functions n = number of [integrative; interactive; bridging] KMS functions implement(i) := 1 if function i is implemented
implement(i) := 0 if function i is not implemented
g(KMS) = group of [integrative; interactive; bridging] KMS functions n = number of [integrative; interactive; bridging] KMS functions frequent(i) := 1 if function i is used “often” or “always”
frequent(i) := 0 if function i is either not implemented or used less frequently
Extension of KMS usage was measured per group of KMS functions—integra- tive, interactive and bridging functions—as the number of implemented functions.
Intensity of KMS usage was also measured per group of KMS functions as the number of frequently used functions (often or always used). These measures will be used for the analysis of the relationship between the usage of KMS functions on the one hand and the achievement of business goals on the other hand122.
KMS functions in organizations were used in a variety of ways. First of all, most organizations used basic functions for information retrieval and communication, like keyword search, navigation, email and email distribution lists. The use of more advanced functions, however, was preserved for a much smaller group of organiza- tions. Integrative KMS functions were used widely and frequently, especially to support knowledge acquisition and publication and to a somewhat lesser extent knowledge organization. Generally, interactive KMS functions were used less fre- quently than integrative KMS functions. Thus, even though the use of interactive functions was up when compared to previous studies, organizations at least until 1999 did not predominantly use communication-oriented functions as was hypoth- esized by Blackler (1995). Hypothesis 5: ’Organizations converge in their use of ICT and increasingly use communication-oriented functions of knowledge man- agement systems’ therefore was not supported.
The reason for this might be attributed to the problem of bandwidth preventing especially multinational organizations from a substantial organizational effort to boost the use of bandwidth-intensive functions, such as videoconferencing or
122. See section 15.2.4 - “Correlations with goals” on page 575.
extension g KMS implement i( ) ---n
i=1 n
= ¦
intensity g KMS frequent i( ) ---n
i=1 n
= ¦
Stt.010.Mssv.BKD002ac.email.ninhddtt@edu.gmail.com.vn.
advanced tele-learning applications like live broadcasting of videos or video server. This might change in the near future, though, as investment in network infrastructure is considerable, though not at the same pace throughout the world.
This means that especially large multinational organizations will encounter band- width problems in certain parts of the world even in the longer run. This is espe- cially true for their operations in countries with infrastructure that is still chal- lenged by the large gap between supply and demand if they try to install KM pro- grams that should discriminate some of their subsidiaries, e.g., in the third world.
Only a small portion of the list of KMS functions could be compared to the more superficial analysis of KMS functions performed in previous studies. The share of organizations that used functions that actually could be compared, did not rise with respect to previous studies. Hypothesis 6: ’Compared to earlier studies signifi- cantly more organizations use ICT in general and knowledge management systems in particular to support their KM activities’ was not supported for KMS functions.
However, according to the interviews it seems that organizations have imple- mented a number of advanced KMS functions only recently so that the use of these functions might well be on the rise.
Table C-49 gives an overview of the results for the different KMS architectures.
The aggregation of means for ordinary scaled variables cannot be interpreted seri- ously. The numbers are meant to give a general indication of which categories of functions were used by what share of organizations and in what frequency123.
The average number of functions per organization was calculated as the mean of the number of integrative, interactive and bridging KMS functions implemented per organization. The average share of organizations with functions was computed using the following formula:
123. The mean over means and the mean over medians differ only marginally, though, and the general picture has been confirmed in the interviews, so that the results all point in the same direction.
TABLE C-49. Aggregated comparison of KMS functions according to KMS architectures
KMS architecture number of func-
tions
average number of KMS functions per organization
average share of organizations with
KMS functions
mean frequency of usage of KMS functions
integrative 28 10.03 (35.82%) 53.04% 3.12
interactive 20 7.21 (36.05%) 50.24% 2.74
bridging integra- tive and interactive
14 4.01 (28.64%) 42.59% 3.00
total 62 21.25 (34.27%) 49.77% 2.97
Stt.010.Mssv.BKD002ac.email.ninhddtt@edu.gmail.com.vn.
avgorg = average share of organizations with KMS functions
g(KMS) = group of [integrative; interactive; bridging] KMS functions kms = number of [integrative; interactive; bridging] KMS functions
nimplement(i) = number of organizations that have implemented KMS function i ntotal(i) = total number of organizations that have responded to the question about KMS function i
The mean frequency of usage of functions was calculated aggregating the mean values for each function to the groups integrative, interactive and bridging KMS functions:
Pfreq = mean frequency with which KMS functions were used
g(KMS) = group of [integrative; interactive; bridging] KMS functions n = number of [integrative; interactive; bridging] KMS functions P(i) = mean frequency with which function i is used
Functions bridging integrative and interactive KMS were used in substantially fewer organizations than functions of the two other groups, integrative and interac- tive KMS functions. Those organizations that actually had implemented bridging functions used them in about the same frequency than functions belonging to the other two groups. In the case of interactive KMS functions, removing the two func- tions email and email distribution lists revealed that the more advanced interactive KMS functions were implemented and especially used substantially less frequently than integrative KMS functions124.
Extension and intensity of usage were also compared for organizations with or without KMS. The adjusted significance level using a Bonferroni type correction was 0.0166125. Table C-50 shows the corresponding statistical results.
As expected, organizations with KMS had more KMS functions implemented and were also using these functions more intensively than organizations without KMS126. With the exception of the extension of interactive functions all correla- tions were significant. Hypothesis 18: ’Organizations with KMS have a larger number of KMS functions than organizations without KMS’ was supported. The
124. The values for interactive KMS without email and email distribution lists were as fol- lows: an average number of KMS functions of 5.70 (31.67%), an average share of 44.68% and a mean frequency of 2.54.
125. See section ‘Statistical analysis.” on page 455 for an explanation of the Bonferroni type correction.
126. All correlations showed the expected negative sign meaning that organizations with KMS were using KMS functions more extensively and more intensively.
avgorg g KMS nimplement( )i ntotal i ---
i=1 kms
= ¦
Pfreqg KMS P( )i ---n
i=1 n
= ¦
Stt.010.Mssv.BKD002ac.email.ninhddtt@edu.gmail.com.vn.
relationships were stronger for integrative and bridging functions than for interac- tive functions. One might conclude that the KMS used in the organizations cur- rently primarily target integrative and interactive functions.
In the following, these results showing the actual use of KMS functions are compared to estimates of the potentials of these technologies or expectancies about KMS functionality. These were studied in the related empirical studies.
In the Fraunhofer Stuttgart study, respondents had to estimate the potentials of modern ICT in the context of KM (Bullinger et al. 1997, 20f). An efficient access to information was considered to carry the highest potentials which would require primarily integrative KMS functions. The support of the identification of experts (yellow pages) was not seen as a promising technology. Respondents believed that their organizations relied on personal networks for the identification of experts rather than on links and electronically accessible information.
Similar results were obtained in the Delphi study. Most respondents of that study (63%) thought the organization of existing corporate knowledge and thus integrative KMS functions were a valuable feature of KMS. A much lower share of organizations (39%) felt that KMS could provide new ways to share tacit knowl- edge and thus require interactive functions (Delphi 1997, 15). However, sharing tacit knowledge was ranked higher in that survey than sharing explicit knowledge (29%) and the support for research and knowledge generation (31%).
In the 2001 KPMG study, organizations also ranked integrative KMS functions as the most important parts of a KMS with an emphasis on knowledge search and publication: development/updating of contents (1.32127), management of existing knowledge (1.48), keyword search (1.62), full text search (1.79). Interactive KMS functions such as determination of contact persons (2.17) or collaboration (2.95) were considered less important (KPMG 2001, 18).
The results of the follow-up study on knowledge management by the FH Cologne performed in 2001 seem to point in the same direction. Respondents esti-
TABLE C-50. Correlations between use of KMS and use of KMS functions aggregated measure for KMS functions Spearman’s rho significance n extension (integrative functions) -.408 0.000555 68 extension (interactive functions) -.284 0.019028 68
extension (bridging functions) -.408 0.000553 68
intensity (integrative functions) -.498 0.000015 68 intensity (interactive functions) -.380 0.001407 68
intensity (bridging functions) -.528 0.000004 68
127. The scale extended from 1—very high importance to 5—very low importance of the function.
Stt.010.Mssv.BKD002ac.email.ninhddtt@edu.gmail.com.vn.
mated the highest potentials for KM tools in the areas integration and consolida- tion of informations (81% said this area is important or very important), exploita- tion of existing information (80%), communication between employees (73%), a higher transparency (68%) and improved collaboration (68%) whereas document management on average was found less important (56%, Dửring-Katerkamp/Tro- jan 2001, 3f). The latter result might already suggest a trend towards the more advanced KMS functions which define traditional document management as a solid basis, but not a focus of KM.
Bullinger et al. conclude, though, that many organizations might underestimate the potentials of KMS to support networks of employees and their sharing of knowledge (Bullinger et al. 1997, 21). One reason might be that advanced func- tions bridging integrative and interactive KMS are comparatively new and not yet well known in the organizations. Thus, organizations might view the simpler inte- grative KMS functions as more promising in the first place whereas a successful application of the more advanced functions bridging integrative and interactive KMS requires a more sophisticated KM approach. Another reason might be that the larger organizations were, the more respondents felt that their IT infrastructure was unsuited for a support of KM (Bullinger et al. 1997, 40). Additionally, larger organizations seemed to suffer more under a lack of visibility of knowledge and knowledge demand and under bureaucracy and hierarchical structures (Bullinger et al. 1997, 40), which might make it even harder for this group of respondents to envision a successful use of KMS.