INTRODUCTION
Technology-using organizations are interested in attempting to mitigate computer gold-bricking by employees. Computer gold-bricking involves the use of the computer for non-business use such as playing games, e-mailing friends, or accessing non-work-related web sites. Given the increasing use of the computer in business, the potential for abuse is also increasing. How serious is this problem?
THE PROBLEM
Computer gold-bricking appears to have begun with the popularity and enormous availability of computer games. Games automatically accompany the Windows operating system and they are given as incentives to subscribe to computer magazines. There are many commercial games and literally hundreds of shareware and freeware games. In addition, major software game manufacturers have Internet sites for games. For example, Microsoft's Intemet Gaming Zone is an online gaming service with 1.1 million registered users and over 6,500 peak simultaneous users (13).
The popular media including the Internet publish various ways to waste time at work including surreptitious gameplaying (2). Don's Boss page (3) lists Worker-Friendly sites for a safe surfing experience that contain a Panic Button as a public service to keep at-work internet time as secret as possible. Hide Your Windows OuttaSight will keep your boss from seeing that you're playing Riven. It will cleverly conceal any active window and password-protect the program so that others can't see what you've hidden (12). It seems evident that there are many people who are engaging in such stealthy activities. In any large IS shop, there are always stories circulating about elaborate methods of computer gold-bricking (4).
It does seem patently obvious that games are the antithesis of business activities. Ambrosia Software, Inc. advertises its lines of Macintosh software as "productivity software and antiproductivity (game) software." GameWarden (6) reports surveys that 44% of workers with computer play games on them in the office wasting 26 million hours of productive work at a cost of $750 million annually. Popular publications note that the nationwide study by SBT Accounting Systems (17) found that U.S. workers play computer games in the workplace. Similarly, references are made to a press release from Emily Coleman in 1995 of Coleman & Associates Inc. reporting the results of a national phone survey of more than 1,000 adults with a margin of error of 3.5%. The study claimed that 23% of computer gamers played their most recent game at the office and that middle management played games most frequently (1, 15, 20).
Unfortunately, the above studies on computer goldbricking are not in the professional literature so their methodology cannot be verified. Citations in the professional literature referring to the above studies were uneven and could not be verified. After a thorough search, it became clear that virtually all of the information on employee computer goldbricking was initially reported in popular business, newspaper, and magazine publications with a similar pattern.
The standard publication on computer gold-bricking begins with an attention-getting example such as employees playing hearts on the network or downloading pornography. The implication is that this behavior is both inappropriate and non-productive. This is followed by one or more striking and unverifiable references on how much it costs to play games or how much time is wasted by employees surfing inappropriate Internet sites. This is followed by more unverifiable and often hearsay evidence on how companies are cracking down by instituting harsh controls, issuing stem warnings, and even firing the gold-brickers. The conclusion is a mandate that someone must take control of the situation before it gets out of hand. Smart Workplace practices (18), Harmon (7) in the NY Times, Eng and Schwartz (4), and http://www.antigame.com provide references to computer-control articles in newspapers and trade publications. GameWarden (6) has an extensive site with such information including a white paper on antiproductive computer gaming, but the information must be taken with a grain of salt because they market AntiGame, a software package to monitor and control the use of computer games by employees.
Though the research itself is marginal, the popular literature does a good job of stating the logical arguments for and against the use of computers for non-business purposes. These arguments are summarized below.
1. It wastes employee productive time.
2. Valuable computer resources are consumed.
3. Games may have been pirated, putting the company in jeopardy.
4. Programs brought in often contain viruses.
5. Such computer use presents a bad image to visitors. On the other hand, some non-business computer activities can be useful for a business, such as playing games (16). Eng and Schwartz (4) note that "some companies actually promote games in the office to reduce stress and increase productivity." Some benefits include:
1. Teaching the new user to become familiar with the system.
2. Desensitizing people who have a fear of computers.
3. Increasing enthusiasm and interest in computer use.
4. Relieving stress and opening blocked creative channels.
5. Promoting camaraderie among workers.
In fact, there is no reasonable evidence that computer goldbricking has had any identifiable effect on productivity. In one of the most thorough studies of the relationship between computers and productivity, Landauer (9) does not suggest that such computer gaming is in any way related to the paradoxically small impact that computers have had on national productivity.
ORGANIZATIONAL POLICIES
The non-business use of computers is not a recent phenomena though popular articles suggest that this is a new and dangerously non-productive phenomena that must be quickly and firmly dealt with (6). Computer gold-bricking has been a problem since the advent of computers. One of the first games on mainframe computers was Adventure, a role-playing game where the player attempts to find treasures in a cave (11). However, organizational policies in the past were ad-hoc and there are very few references to control of non-business use of computers in the professional literature prior to 1990. Of course, the exponential increase in the use of computers in business has changed the situation. Rather than a few mainframe managers concerned with inappropriate gaming, now there are many more individuals involved in similar activities. Even if the proportion of non-business computer use has not significantly changed, there will be substantially more people involved. This is an important point since there is no evidence to suggest that the proportion of computer goldbricking has significantly changed.
Although mainframe managers prior to the 1990s did have to deal with computer gold-bricking, there is no research to tell us the significance of the problem or the standard methods of resolution. Today, the media informs us of newsworthy methods of dealing with the situation.
The Illinois Governor ordered the removal of Solitaire and Minesweeper, games that come with the Windows operating system, from about 10,000 Illinois state computers (5). The only reason given was that such games were not appropriate. The directive did not come as a result of excessive game playing by state workers nor did the article state that there were any complaints.
"Corporations are quickly realizing that perils that accompany the benefits of giving employees Internet access" (14). We must stop them with filtering and monitoring software designed for the corporate market including W3 (21), Surfcontrol Scout, Sentinel Policy, and Cyber patrol Corporate. These packages monitor and restrict Internet access. Some include a stealth mode that quietly stores the violator's ID, time, date, application, and screen images in a database for later review by the boss. "This will give the boss the evidence to fry a lying employee who won't admit illegal Internet access" (21).
AntiGame (5) can detect and optionally remove games from a business network's server. Its database includes thousands of popular games. It can monitor, terminate or (stealth mode) save for later report. Games can be allowed or disallowed by selected hours, by specified PCs, or by total time allowed per day. You (the boss) can `blow him away, challenge him to a 6 p.m. Deathmatch, or just let him know you're there, watching and waiting" (6).
The above examples follow a logical, but ill-fated scenario in attempting to control the bad office employee(s). First, a manager/supervisor/boss becomes aware of a significant employee-based problem. The employees are told to quit doing the problem activities. If the problem is not solved at this level, stronger measures are implemented. In most cases, this process continues until the costs of the stronger measures exceed the cost of the original problem. Ironically, the control costs are no longer an issue. The issue has resolved to one of control rather than to one of least cost. While the problem is apparently solved, these issues remain:
The costs to control the problem are greater than the original losses of the problem.
Morale drops with the philosophy of punish everyone for the misdeeds of a few.
There is an uneasy feeling that some employees are still evading the controls.
METHODOLOGICAL CONSIDERATIONS
Due to the nature of the research topic, traditional research methodologies are not appropriate. For example, it is highly unlikely that any survey would reveal valid answers from employees concerning how much time they are spending playing games or accessing inappropriate Internet sites. SBT's study (17) noted that "employees were reluctant to disclose the amount of time spent playing computer games on office computers." Employees are unlikely to own up to computer gold-bricking since the reaction from business has been a wellpublicized censuring those who engage in it.
Because of self-interest, consulting firms selling advice and software to help control computer gold-bricking would be a poor choice for valid information. Similarly, managers and executives of computer divisions, although useful for anecdotal information, can only focus on the perceived troublemakers in their local situation so that a representative picture of computer gold-bricking is not reasonably possible from this group.
Given these methodological constraints, an appropriate step would be to determine the attitudes toward computer goldbricking according to standard demographic variables. Accordingly, for this research a survey instrument was constructed to provide baseline results on attitudes toward the non-business use of computers in the workplace with the intention that it could be used profitably by businesses. The results of an attitudinal survey on computer gold-bricking may also provide direction in formulating organizational policies for effectively dealing with the situation.
RESEARCH AND METHODOLOGY
To explore the problem, we developed a Likert attitude scale using the good and bad perspectives of computer use for non-business purposes developed from the literature. With college students as the sample, we included demographic items for class rank, major, gender, and computer experience, and the scale was pilot tested and slightly revised. The final survey instrument is shown in the appendix. The items were scored from I (SA) to 5 (SD) except for items 4, 6, and 10-13 which were scored opposite to reflect the converse wording. The total score was a sum of the scores on the individual items and intended to reflect the respondent's attitude toward playing games and Internet use in a business situation. The higher the total score, the less favorable the attitude toward computer games and Internet access in the workplace.
The survey was administered to a selection of classes taking some course in Information Systems with students enrolled from across the campus of a Midwestern university. Seventy-eight (78) usable surveys were obtained from five different courses representative of the campus population. The distribution of total scores was relatively normal with a mean total respondent score of 39.3, variance of 35.6, and a range from 27 to 57. For the item analysis, the total scores by respondents on the items were used to divide the sample into low and high groups based on a median split. The group mean scores on each of the items were then compared using a standard t-test.
The item analysis in Table I shows that the survey results are consistent with an internally reliable instrument. All of the items except #10 showed a discrimination probability less than .025 with most exceeding the .001 level. Therefore, item 10 was dropped from the statistical analysis and will be discussed in a later section. Dropping item 10 did not affect the item's ttest scores, but the total item average dropped to 2.708 while the test reliability coefficient improved to .751. The test reliability coefficient was calculated using the KuderRichardson Formula 20 as an estimate of the reliability of the scale. It is often interpreted as the average of all possible splithalf correlations of the scale (8, 19).
Table 2 shows the demographics of the sample. All class ranks were represented with an average rank of slightly less than a Junior. Males and females were split approximately 40/60 which is not significantly different from the campus population. All levels of computer experience were reported with an average of "capable" experience.
Attitude Scale Results
There were no statistically significant differences among the attitude scale total scores by major or by gender. The high multicollinearity between rank and computer experience (R=.52), P<.0001) was resolved by the multiple regression showing that computer experience was the more important correlate of attitude scale score. The highly significant results were
Attitude = 16.23 + 5.4* [Computer Experience], R = .58, F = 38.8, MSE = 71.0, P < .0001
This means that the more computer experience a person has, the more likely they will feel that game playing and nonwork-related Internet access on the job are not appropriate and should be controlled.
CONCLUSIONS AND RECOMMENDATIONS
The Item 10 Anomaly
Item 10 shown below was dropped from the attitude scale because it did not discriminate.
10. Office workers should not play computer games if the public they serve can see them playing the games.
This was the highest scored item with virtually everyone agreeing (mean = 4.128, Strongly Agree = 5). We concluded that no one approves of this activity. Therefore, businesses should ensure that employees do not play games in view of their public.
Is Employee Control Really Necessary?
If our results would be confirmed in an organization, i.e., that the more computer literate employees are in favor of restricting game playing and Internet access, then there should be no serious problem in that organization. It is very likely that the normal social processes in the informal organization will exert sufficient control to keep most people in line.
With our findings, we suggest that organizations should be particularly careful of the punitive approach so often advocated in the popular press. There are more proactive, constructive approaches to dealing with computer gold-bricking (15).
Restrict the non-business use of computers.
Provide clear guidance to employees on management policies.
Support acceptable games and Internet sites.
Designate some computers for game-playing. Restricting the non-business use of computers may be viewed as a punitive measure. Any policy involving general restrictions is based on mistrust and deepens the employeremployee adversarial relationship unnecessarily (1, 10, 20). The more moderate course of action is to follow the last three points with emphasis on a reasonable written policy providing clear guidance of acceptable activities. Our findings suggest that most employees will agree and abide by such policies. Using software or talented employees to catch computer goldbrickers is not recommended due to the costs and the probability of lowering morale with little evidence that the gold-bricker minority will be stopped. In fact, those people may see any company attempt to restrict them as a challenge.
The Gold-Bricker Minority
It is likely that the gold-bricker minority is responsible for much of the publicity and management attempts to control this problem. We would estimate that this group has not significantly changed over the computer era remaining at I or 2% of all computer users (11). Our research on this topic indicates that some organizations will have some highly computer literate employees who do play games and access non-business sites on the internet.
An information systems analyst (who required anonymity) at a large corporation who was a FreeCell addict stated to us that the played about ten games a day at about five minutes per day. This is a significant amount of time, almost one hour out of an eight hour workday. The posted company policy prohibited games of any type, so the analyst, a 20-year veteran, wrote a program that encrypts and saves the FreeCell files. In the morning, he runs the program to unload and decrypt the files into a workspace where all of the other daily work is located. At the end of the day, the FreeCell files are duly encrypted, renamed, and saved so that they do not appear on the system. The saga here is much more elaborate than described, so consider the time that this analyst has spent in outwitting the No Games directive. The preceding scenario suggests that trying to stop the determined gold-bricking minority is not only futile, it can be very costly. Though this is a bitter pill for some managers, we suggest that the goldbricking minority be ignored.
If an organization wishes to replicate our research. we caution them to avoid the validity problems mentioned earlier by having the survey done by an outside firm that can guarantee the anonymity of the respondents. Then an appropriate modification of the survey instrument shown in the Appendix can be used to benchmark attitudes.
No comments:
Post a Comment