Managing WWW in Public Administration: Uses and Misuses

Abstract

The Internet has emerged as one of the most prevalent forms of communication media in and among public organizations. The construction and management of The World Wide Web (Web) sites are becoming essential elements of modern public administration.This study is intended to provide an in-depth evaluation on the Web sites of Taiwan's central government based on the Web performance indicators provided by Nielsen (2000).Based on the Nielsen's indicators, the authors carefully studied and coded each individual Web site of the central governmental agencies.The coding results indicate that the government Web sites in general have made many of the mistakes as predicted by Nielsen.Most of the agencies need to improve the coordination between Web designers and the line mangers of the agencies.In light of these research findings, this paper provides a number of strategies to improve the Web design practices in the public organizations.

Introduction

The Internet has emerged as one of the most prevalent communication media in and among public organizations.Various surveys have documented the rapid diffusion of the Web in public administration and signaled the profound changes taking place in public managers' work settings (e.g. http://www.cyprg.azrizona.edu).While the construction and management of Web sites are becoming essential elements of modern public administration, little is known about the effectiveness of the public Web sites.Given the substantial investment in time and other resources being committed to the governmental on-line efforts, it is essential to begin undertaking the evaluation of governmental agencies' Web sites in terms of quality and effectiveness.

To remedy the lack of evaluation research on the governmental Web sites, we conducted a national study on the public Web sites of the central government in Taiwan.The central government of Taiwan consists of thirty-three ministry level agencies in charge of the national affairs of the various policy areas.We chose to study at the level of the central government for several reasons, including the early development of WWW sites at the central governmental agencies, the relative diversity of the ministry functions and organizations, and their tractability as units of analysis when compared to the lower levels of the government.

Following the reinventing government movement in the U.S. (Gore, 1994; Osborne & Gaebler, 1992), the central government of Taiwan launched its own large-scale administrative reform aimed at modernizing the governmental services soon after the reelection of the President Lee Teng-Hui in 1996.A major theme of the reform effort has been the revamping of the central government's management information system.The Prime Minister asserted that he was going to establish an "electronic government," and the first step was to help the government agencies go on-line (Wei, 1999).By the time we conducted this study, this reform had been underway for about four years.While all of the central governmental agencies have established their own publicly accessible Web sites for at least two years, there has been virtually no systematic evaluation to justify these on-line efforts.The increasing transparency of the Internet surely has led to questions of its impact on the administrative ethics of the public employees (Menzel, 1998) as well as on the personal and organizational behaviors of the citizens (e.g. Hsian, 1999). However, there is a lack of evaluation about the contents of governmental Web sites.

The study reported in this paper conducted structured content analysis on the governmental Web sites in the context of Web usability. The objectives of this study were to provide useful management recommendations in terms of Web site design, and to suggest a systematic way for assessing public Web sites in general.

Research Conceptualization

The potential of the Web as a new public medium mainly depends on the content of Web pages.Research on the content of the Web has progressed in two directions.In addition to the debates around the regulation of the Web content (that is, what information can/should be put on the Web), and the functions included in public Web sites (Nunn & Rubleske, 1998; Scavo & Shi, 1999), many researchers have begun to ask questions about how the Web presents information.More specifically, the focus has moved from the types and meanings of the information contained on a Web site to include the way in which the content is presented and its implications for Web usability.Mitra and Cohen (1999) argue that this critical textual analysis is to move beyond an analysis of the volume of text to the level of understanding the effectiveness of the organization and presentation of specific information; they offer interpretations of how a text can become meaningful and effective when placed in the cyberspace.

Characteristics of the Web

A critical textual content analysis of Web sites requires a clear understanding of the unique characteristics of the Web.The Web can be considered as a multifaceted mass medium that contains many different configurations of communication (Morris & Ogan, 1996). The glue that holds the modern Internet world together is the Web programming language, namely HTML (Hypertext Markup Language).The term "Hypertext" was first coined by Theodor Holme Nelson, a recognized ideologist of Hypertext, in reference to a radically new way of storing and viewing information.Instead of gathering or retrieving information sequentially, information recorded with Hypertext is fashioned in multiple layers.An automated index is built into the Web document.The intertextuality and non-linearity of HTML enable Web pages to connect various virtual contents with specific "links" which allow on-line users to move among points and "nodes" (Howell, 1992).

Therefore, Web pages are more than printed pages posted electronically.The Web offers many new opportunities as well as challenges to modern organizations (Parker, 1997; Mitra & Cohen, 1999).First of all, the Web makes it easy for transmitting information in a timely fashion.Changes to a Web site can be published in a relatively short time when compared to the redesigning, production and distribution processes of most printed media.Secondly, Web pages can include larger amount and more variety of information without incurring major printing and distribution costs.On the Web, costs do not necessarily increase as the amount of information being communicated increases.Furthermore, multimedia objects including drawing, photograph, animation, sound, video, and computer applications, can be incorporated into Web pages at a low cost to enhance the Web's communication effects.

On the other hand, as Lynch and Horton (1999) pointed out, the originators of the Web intended the Web to be a device-independent method for exchanging documents.With their focus on structuring and sharing documents, they ignored the visual logic or graphic design aspects of Web information delivery that are critical to effective communication.One Web characteristic that sets Web development apart from traditional media design is the element of lack of control.Unlike designers of printed media, a Web designer somewhat loses control over how on-line users will browse the pages, the appearance of the fonts and colors used on a page, and the sizes, proportions and exact locations of the different Web texts.On the Web, users largely determine their own navigation paths, and they are free to "jump" to any location that interests them.In addition, designers cannot know what exact computer equipments the various potential users have, or what fonts and software have been installed in the users' computers.The exact way WWW pages present information would be partly determined by the users' own environment.Parker (1997) indicates that while Web pages look in many ways like printed pages, the Web is a new medium, and hence it requires a new approach to develop a successful Web site.

Usability of the WWW

The process for developing a WWW site generally consists of three stages (e.g. Giudice and Goodman, 1999; Davis & Merritt, 1998):

I)defining the mission, goals, potential users and general strategies of the site;

II)designing and choosing the appropriate unified style, information architecture, technologies, and navigation paths; and

III)producing, integrating, testing, and refining the programming, typography, editorial style, graphic design, and multimedia objects of a WWW site.

Ideally, the Web development process should be a user-centered process.It is during each of these stages that the Web designer takes the needs of the potential users into consideration, and hence builds in the usability of the site.

While the Web has its unique characteristics compared to the printed media, many of those who are responsible for developing their organizational Web sites apparently have not paid enough attention to the distinctions between them. According to a research report released by Creative Good, Inc. (Speigel, 1999), key errors in Web design could end up costing online retailers as much as $6 billion (US$) in unrealized revenue in a holiday season.The report showed that 39 percent of test shoppers failed in their buying attempts because sites were too difficult to navigate.Additionally, 56 percent of search attempts failed.Furthermore, the report indicates that a dollar spent on advertising during the 1998 holiday season produced $5 in total revenue, while a dollar spent on customer experience improvements yielded more than $60.Finally, the report concluded: "the e-commerce winners will be those sites that create a good shopping experience for their customers. Quick, simple, clear sites will enjoy higher conversion rates than flashy or slow sites." (p. 1).

Rosenfeld and Morville (1998) argued that most of the successful Web sites share the same characteristics: they integrate each of the key aspects of Web site development including information architecture, technical design, and graphic design.They summarize the common reasons why users like a particular Web site as aesthetics, big ideas, utility, "findability", and personalization.In short, the success of Web site mainly depends on the usability of the site.

Evaluating Web Usability

Instone (1997) agrees that the most important aspect of Web design is usability.To improve usability, Instone suggests web operators employ a heuristic evaluation originated by Jakob Nielsen in the early 1990s be undertaken to make a Web site more usable.Heuristic evaluation involves the study of a user interface by a small set of evaluators who look for violations of common usability principles.The problems identified in a heuristic evaluation could be tackled right away, leading to a redesign, or used to focus a usability test.Heuristic evaluation is well suited for the Web because it can be easy, fast and inexpensive. Basically, it involves identifying the chosen heuristics (performance indicators), gathering opinions about the usability of particular sites, merging and rating the problems that were identified, and then trying to work toward solutions.

The Web literature has provided a comprehensive list of performance indicators that can be used to evaluate the usability of Web sites. Rosenfeld and Morville (1998) argue that Web site architects must strive to learn what it is like to really use the Web on a regular basis to reach a high level of user-centered awareness. It is the kind of awareness that distinguishes successful Web designers from the others.They summarize seven reasons why most people dislike a Web site:

1. Difficulty of finding the needed information;

2. Poor graphic design and layout;

3. Gratuitous use of animations and other Web technologies;

4. Inappropriate tone;

5. Designer centeredness;

6. Under construction; and

7. Lack of attention to detail.

In a similar vein, Nielsen (1996) warns that while maintaining Web sites has increasingly become an integral part of organizational management, critical assessments are needed to avoid mistakes at all levels.Nielsen has compiled the most comprehensive set of Web usability criteria.He first identified the top ten mistakes of Web design by analyzing the then existing Web sites in 1996.Three years later, he revisited the ten mistakes and found that the top ten mistakes in Web design that he identified in 1996 were still surprisingly common, and are listed below:

1. Using Frames;

2. Gratuitous Use of Bleeding- Edge Technology;

3. Constantly Running Animations;

4. Complex URLs (Uniform Resource Locators);

5. Orphan Pages;

6. Long Scrolling Pages;

7. Lack of Navigation Support;

8. Non-Standard Link Colors;

9. Outdated Information; and

10. Overly Long Download Times.

Furthermore, according to Nielsen's observations, new Web technology and new applications for the Web have introduced a new list of design mistakes (Nielsen, 1999).The ten new mistakes are as follows:

1. Breaking or Slowing Down the Back Button;

2. Opening New Browser Windows;

3. Non-Standard Use of GUI Widgets;

4. Lack of Biographies;

5. Lack of Archives;

6. Moving Pages to New URLs;

7. Headlines That Make No Sense Out of Context;

8. Jumping at the Latest Internet Buzzword;

9. Slow Server Response Times; and

10.Advertisement-like Design.

Many of the Web design and management problems are almost inevitable for those who are new to the Web project development, unless explicit action is taken to avoid them.

Research Methods

This study is intended to provide a usability evaluation on the Web sites of Taiwan's central governmental agencies based on the Web performance indicators provided in the literature.This research uses a form of the case survey method advocated by Yin and Heald (1975) to assess the degree to which that each of the Web sites committed the Web design mistakes.Web improvement recommendations are then made based on the collection and analysis of the identified design problems.

The authors organized a panel of coders composed of sixty college students who were taking the course of "Administrative Information Management" during the period of undertaking this research.These students conducted structured content evaluation on each of the existing thirty-three Web sites of the national level executive agencies.

To establish the framework for analyzing the public agencies' Web sites, eighteen Web design and management mistakes were selected from the works of Rosenfeld and Morville (1998) and Nielsen (1996 and 1999).It should be noted that while almost all of their indicators are applicable to evaluating various Web sites, only the indicators whose relevant dimensions can be captured on the Internet were selected for this study.For example, while we recognize that "lack of archives" and "moving pages to new URLs" could be two major mistakes of Web sites, we had to abandon them since they cannot be measured by analyzing the content of the current Web sites and there was no historical data on the two dimensions.As shown in Table 1, the eighteen Web indicators are further classified into 3 groups according to their influence on the different Web development stages, namely defining, designing and producing, using the concept of Giudice and Goodman (1999).

(Table 1 about here)

Each individual Web site was then carefully studied and coded on the 18 indicators.The coders were asked to rate the assigned public Web sites on each of the selected dimensions using a four-point Likert scale, with 1 being "no problem on this dimension", 2 being "minor problem", 3 being "medium problem", and 4 being "serious problem identified".In other words, a perfect Web site would be scored as 18 and the most problematic would be scored as a 72.

T
To enhance the validity of the study, all of the coders were given specific instructions and tested on the coding criteria by the authors two weeks before conducting the coding.In addition, to ensure inter-rater reliability, each of the Web sites was coded in a double-blind fashion by two subgroups.Whenever an unmatched result was found, the authors asked the two sub-groups of coders to discuss the reasons of their choices, and then to make a collective decision on their ratings.

Results

The results of analyzing the exiting 33 Web sites of Taiwan's central governmental agencies based on the eighteen selected indicators are presented in Table 2. No agency received a perfect score of 18 (scored as 1― "no problem"― on all of the 18 indicators).The highest score of 30 was achieved by the Central Personnel Administration.Only one-third of the agencies (11) achieved a score lower than 36 (mean = 2, "minor problem").Among the agencies that received the best scores are the Mainland Affairs Council (31), the Fair Trade Commission (32), the National Palace Museum (33) and the Department of Health (33).Six agencies scored 34 or 35, and the remaining 22 agency web sites scored no more than 43.

(Table 2 about here)

In addition, Table 3 through 5 list the indicators that comprise each of the three Web development stages. These tables also give the percentages as well as the overall means of the Web mistakes identified in the Web sites.The most common problems were lack of navigation support, design inconsistency, overly long download time, lack of foreign language version, and orphan pages.

Stage I: Defining

The "Defining" stage includes three indicators.Each of these items is viewed as a negative contributor to a Web site's usability.As Table 3 shows, the item that received the highest score in the table is "lack of foreign language version." One of the unique features of the Web is its freedom from expenses related to quantity, quality and, most importantly, distance of communication.Since the Web sites we analyze in this research represent many national governmental functions, it should be expected that people speaking different languages would seek information about Taiwan from the sites.

(Table 3 about here)

The next most serious problem is "designer centeredness."Twenty of the Web sites were found to have minor to medium level problems on this indicator.Four of the sites are scored as 4, indicating very serious problems observed on this dimension.Web sites having this mistake are more internally focused.A common characteristic is having home pages full of mission statements, photos of the senior managers, and history of the organization.

Finally, the problem of "mirroring the organizational chart" is one of the identified indicators that the governmental Web sites analyzed here generally perform better. Among the thirty-three Web sites analyzed in this study, twenty-four sites (72.7%) receive a score of 1, indicating no problem on the dimension.

Stage II: Designing

Table 4 lists the five items that make up the "Designing" stage.One notable item in the table is the item on navigation support. The strong positive mean of 3.0 shows the high degree to which this problem was identified.All of the central governmental Web sites are found to have a certain degree of this problem, while one-third (33.3%) of them are particular serious in which neither a navigation guide nor a search function exists.

(Table 4 about here)

A related mistake, "orphan pages", is also found on most of the central governmental Web sites.More than 80% (81.8%) of the Web sites contain one or more pages that include no indication of what Web site they belong to.

Similarly, among the surveyed Web sites, two-thirds (66.7%) use frames on their Web pages.Splitting a Web page into frames is very confusing for users because frames break the unified model of the Web and introduce a new way of looking at data that has not been well integrated into the other aspects of the Web.While half of the Web sites that use frames (33.3%) provide a more consistent browsing path (scored as 2 = minor problem), the others tend to make their hyperlinks on the home page unpredictable.

On the other hand, only about ten percent (9.1%) of the governmental Web sites analyzed in this study are found to show "gratuitous use of bleeding- edge technology."We suspect that it is partly due to the fact that the governmental agencies have not utilized the Web very extensively. Scavo & Shi (1999) conducted a survey on 145 municipalities and county government Web sites and found that much of the governmental presence in cyberspace "is no more than an elaborated telephone directory…… Far fewer local governments have taken the additional steps to take fuller advatage of Web technology, by making their Web pages interactive, for example." (p. 259).This observation also seems to be fair for describing the central governmental agencies' Web sites based on our own evaluation of the sites.

Stage III: Producing

The "Producing" stage is made up of ten indicators.The item of "interface design inconsistency" received the highest score in Table 5.Only four (12.1%) of the sites studied here present a generally consistent appearance throughout the sites.On the other hand, as many as one-third (33.3%) of the Web sites were coded as showing serious problems on interface design consistency.

(Table 5 about here)

The second most serious problem in this stage is "overly long download time."Nielsen (1997) argues that ten seconds is the maximum response time before users lose their interest in browsing a new Web page.In this study, the coders were asked to calculate the time needed to download the homepages of the Web sites they were assigned, using a regular telephone dial-up connection during the day time.Download time of a home page was coded as “1” (no problem) if it is under 10 seconds; “2” if it is between 10 and 20 seconds; “3” if it is between 20 and 30 seconds; and “4” if the download time is over 30 seconds.The results show that only 2 (6.1%) of the agencies’ sites were coded as 1; seventeen (51.5%) sites were coded as 2; eight (24.2%) sites were coded as 3; and the other six (18.2%) sites were coded as 4.

The other items that received a mean higher than 2 include "lack of biographies," "long scrolling pages," "non-standard link colors," "constantly running animations," and "outdated information."With respect to the item on biographies, only five (15.2%) of the Web sites provided appropriate biographies of the documents or data presented on their Web sites.The other Web sites either indicated their data sources occasionally or did not contain any biography at all.

Long scrolling pages and non-standard links colors are also common problems identified in this study.Twenty-seven (81.8%) of the central governmental agencies' Web sites have a home page that is longer than what a 17- inch-monitor using 800X600 proportional size could contain at a time. Meanwhile, non-standard link colors are observed on more than 80% (84.8%) of the governmental Web sites analyzed in this study.

About 70% of the Web sites showed some form of constantly running animation on their Web pages.Moving images have an overpowering effect on the human peripheral vision.It is very difficult for the user to concentrate on reading text in the middle of a Web page if there is a running animation hanging up in the corner.

Finally, failing to regularly update the information included in the Web site is a major mistake found among the Web sites.This mistake is observed on two-thirds (66.6%) of the sites analyzed in this study. The problem is especially apparent on two of the Web sites (Council for Cultural Affairs and Council for Consumer Protection) in which some activity announcements over six months ago are still posted while no recent activities are being reported.

Web Design Recommendations

The coding results indicate that most of the Web sites in general have made many of the mistakes as predicted in the Web literature.While there is an integrated measure of usability for each of the governmental Web sites (see Table 2), the Web performance indicators presented in this paper are best seen as elements of a multi-dimensional analysis framework.The purpose of this research is not to seek a single grade for each of the governmental Web sites but rather to develop a set of observations about the individual sites that can be used as bases for future improvement.

Below we propose some design recommendations we think are most important for those seeking to improve the governmental Web sites based on the findings of this research.One could also extract more design implications from the items listed in Tables 3 through 5.

1. Establish a Web Strategic Planning Team

In light of the research findings, the most common mistakes observed on the Web sites of Taiwan government are lack of navigation support, design inconsistency, overly long download times, lacking a foreign language version, orphan pages and lack of biographies.Many of the above mistakes are interrelated and symbolize the failures of coordination between the different phases of the Web site development.

To ameliorate these problems, a formal strategic planning system should be adopted in the agencies that are serious about the electronic governance efforts.Strategic planning has been shown to lead to better development of e-commerce models as well as Web sites (Harmon et al., 2001; Davenport & Prusak, 1997; Goodstein et al., 1993).Successful development of large Web sites usually requires a cross-functional and interdisciplinary team approach. A formal strategic planning on the Web development process can be used to align the positions of the many stakeholders and to improve the coordination between the Web site producers and the line management functions of the agencies.

2. Design for the Users, not for the Senior Management

Designer-centeredness is one of the most serious problems identified in this study.Web designers should keep in mind that the Web sites of public agencies are designed for use by the general public. The Web site should be designed with the users' needs in mind and not to promote grandiose ideas of self-importance. More specifically, some essential questions must be answered at the beginning of developing a Web site:

·Why is this Web site needed?

·Who will it need to communicate with?

·What do you want visitors to the site to take away with them? and

·What valuable information or data does this agency have that it can communicate via the Web site?

3. Take International Audience into Consideration

An organizational Web site can be established as an international communication center readily when there are potential international users.However, in order to function internationally, a Web site needs to communicate in the language(s) that the potential users understand (Galdo & Nielsen, 1996).

According to our study results, few of the governmental Web sites in Taiwan have taken their potential international users into consideration.As the Web becomes more and more international, the designers of governmental Web sites should pay more attention to the needs of international users to take full advantage of the unprecedented international opportunity afforded by the Web.

4. Create Good URLs

The results of this study show that about one third of the Web sites we analyzed contained complex URLs.While most sites now have navigation support, users tend to rely less on URLs to tell them about their locations on a Web site, and long URLs still cause problems.For example, when users email page recommendations to each other, complex URLs could become unclickable and are difficult to directly type in.Therefore, it is imperative that the URLs of a Web site are as easy to remember and to type as possible.In short, the characteristics of successful Web URLs include: 1) Short and simple; 2) descriptive; and 3) memorable (Parker, 1997).

5. Provide Unified Navigation Interface

With respect to user interface, almost all of the Web sites analyzed in this study did not provide adequate navigation support to users.Furthermore, over eighty percent of the Web sites used non-standard link colors, inconsistent interfaces, or both.Web users need support in the form of a strong sense of structure and place.An essential principle in communicating the structure of a Web site to users is keeping all similar tools used for navigating and accessing information consistent.Web users tend to lose their patience with a site when moving between pages on the site and finding drastically varying interface designs (Davis & Merritt, 1998).

Consistency is the key to usable interaction design. When all graphical user interface elements look and function the same, online users are able to build their confidence in using the site more easily (Lynch & Horton, 1999).At least, every page should have a link up to the home page as well as some indication of where they fit within the information architecture of the Web site.In addition, a good search function and a site map to let users know where they are and where they can go can also be very helpful.

6. Do not interfere with the Normal Way Users Browse the Web

Most popular Web browsers stick to certain interface conventions to provide navigation aids to users.Breaching these conventions confuse users and cause browsing difficulties, but this study shows some government agencies do not consistently use interface conventions.

For examples, some of the Web designers intentionally changed the link colors of their sites possibly to provide a better contrast on the screen.However, most browsers specify that links to pages that have not been seen by the user are blue; links to previously seen pages are purple or red.Changing these colors can confuse users since the ability to understand what links have been followed is one of the few standard navigational aids in most web browsers.

In addition, many Web designers open new browser windows for viewing their Web pages, on the theory that it would keep users on the sites.Unfortunately, this strategy disables the "Back"The "Back" button as "the lifeline of the Web user and one of the most used Web navigation feature" (p. 1).Users feel more secure if they know that they can try anything on the Web and alwaysbe saved by a click or two on Back to return them to a familiar location. button.Using the "Back" button is the normal way users return to previous sites.Nielsen (1999) called

7. Write Good Web Text

While the graphic elements of a Web site are important, text remains the core of most Web sites.The Web is a very different medium from print and television and therefore demands a kind of writing that is different from other media (Kilian, 1999). Nevertheless, the data in this study show that the print and television habits influence the way that many Web designers respond to text on the Web.

The Web tends to have an isolating effect due to its long-distance and anonymous nature.As a result, Web users are becoming more skeptical of the information they find online.They want to know who the people behind information on the Web are.In particular, listing an author’s credentials and including citations and references increase trust and help make a Web site less impersonal (Fogg, et al., 2001).

Furthermore, due to the fact that many headlines of WWW pages would be read out of the context of the full page (e.g. a search engine result), the WWW headlines need to be invitingly brief and understandable in one quick glance (Kilian, 1999).More specifically, Web text needs to be plain and meet two goals:

a.Letting users know exactly what information is presented at the other end of the link with no guesswork required; and

b. Preventing users from following the link if they are not interested in the destination page (Nielsen, 1999).

Conclusion

This research not only provides a systematic evaluation on the existing governmental Web sites in Taiwan, but more importantly, offers a useful guide for those who are interested in undertaking Web application development in public management to study the governmental Web sites.

Due to the unique nature of the Web, those who are new to the Web are likely to commit many of the same mistakes as those described in this paper, unless they learn from others' experience.We agree that the main reason that Web design mistakes happen so often is that their designers generally do not use the sites they design on a regular basis (Argus Associates, http://argus-inc.com/design/ Webarch.html; Rosenfeld & Morville, 1998).It is suggested that both constant surveillance of Web sites, and periodic, more systematic evaluation of the Web defining, designing and producing processes to be undertaken to maintain the usability of a Web site.Web designers have to examine their works from the users' point of view.While there is no standard procedure for developing a successful Web site, public managers can start learning the appropriate design practices by avoiding the common mistakes Web designers usually make.

Tables

Table 1: Web Assessment Indicators

Stage I. Defining 1. Designing for the Senior Managers (Designer Centeredness) 2. Mirroring the Organizational Chart 3. Lacking Foreign Language Version
Stage II. Designing 1. Using Frames 2. Gratuitous Use of Bleeding- Edge Technology 3. Complex URLs 4. Orphan Pages 5. Lack of Navigation Support 6. Design Inconsistency
Stage III. Producing 1. Constantly Running Animations 2. Long Scrolling Pages 3. Non-Standard Link Colors 4. Outdated Information: 5. Overly Long Download Times 6. Breaking or Slowing Down the Back Button: 7. Opening New Browser Windows 8. Lack of Biographies 9. Headlines That Make No Sense Out of Context 10.Interface design inconsistency

Table 2: Analysis Results

(1) Designing for the Senior Managers

(2) Mirroring the organizational chart

(3) Lacking foreign language version

(4) Using Frames

(5) Gratuitous use of bleeding- edge technology

(6) Complex URLs

(7) Orphan pages

(8) Lack of navigation support

(9) Constantly running animations

(10) Long scrolling pages

(11) Non-standard link colors

(12) Outdated Information

(13) Overly long download time

(14) Breaking or slowing down the Back button

(15) Opening new windows

(16) Lack of biographies

(17) Headlines that make no sense

(18) Interface design inconsistency









Table 3: Mistakes of Defining Stage

Performance Indicators	No Problem	Minor Problem	Medium Problem	Serious Problem	Mean
Defining
Designer Centeredness	27.3%	39.4%	21.2%	12.1%	2.18
Mirroring the Organizational Chart	72.7%	18.2%	6.1%	3.0%	1.39
Lack of Foreign Language Version	18.2%	30.3%	30.3%	21.2%	2.55

Table 4: Mistakes of Designing Stage

Performance Indicators	No Problem	Minor Problem	Medium Problem	Serious Problem	Mean
Designing
Using Frames	33.3%	33.3%	24.2%	9.1%	2.09
Gratuitous Use of Bleeding- Edge Technology	90.9%	6.1%	3.0%		1.12
Complex URLs	66.7%	24.2%	6.1%	3.0%	1.45
Orphan Pages	18.2%	36.4%	24.2	21.2%	2.48
Lack of Navigation Support	0%	33.3%	33.3%	33.3%	3.00

Table 5: Mistakes of Producing Stage

Performance Indicators	No Problem	Minor Problem	Medium Problem	Serious Problem	Mean
Producing
Constantly Running Animations	30.3%	33.3%	30.3%	6.1%	2.12
Long Scrolling Pages	18.2%	48.5%	27.3%	6.1%	2.21
Non-Standard Link Colors	15.2%	51.5%	30.3%	3.0%	2.21
Outdated Information:	33.3%	39.4%	21.2%	6.1%	2.00
Overly Long Download Times	6.1%	51.5%	24.2%	18.2%	2.55
Breaking or Slowing Down the Back Button:	75.8%	6.1%	12.1%	6.1%	1.48
Opening New Browser Windows	66.7%	12.1%	9.1%	12.1%	1.67
Lack of Biographies	15.2%	48.5%	33.3%	3.0%	2.24
Headlines that Make No Sense Out of Context	36.4%	42.4%	15.2%	6.1%	1.91
Interface Design Inconsistency	12.1%	24.2%	30.3%	33.3%	2.85

Notes and References

Applehans, W., Globe, A. & Laugero, G. Managing Knowledge: A Practical Web-Based Approach, Berkeley, California: Addison-Wesley, 1999.

Brand, S. How Buildings Learn: What Happens After They're Built,New York: Viking, 1994.

Choo, C. W. Information Management for the Intelligent Organization : The Art of Scanning the Environment, Medford, N.J.: Information Today, 1995.

Davis, J. & Merritt, S. The Web Design Wow! Book: Showcasing the Best of On-Screen Communication, Berkeley, CA: Peachpit Press, 1998.

Davenport, T. H. & Prusak, L. Information Ecology: Mastering the Information and Knowledge Environment, New York: Oxford University Press, 1997.

Fogg, B., Marshall, J., Laraki, O., Osipovich, A., Varma, C., Fang, N., Paul, J., Rangnekar, A., Shon, J., Swani, P., Treinen, M. "What Makes Web Sites Credible? A Report on a Large Quantitative Study," CHI, 3(1): 61-68, 2001.

Galdo, E. D. & Nielsen, J. (ed.) International User Interfaces, New York: John Wiley & Sons, 1996.

Goodstein, L., Nolan T. & Pfeiffer, J. W. Applied Strategic Planning, New York: McGraw-Hill, Inc., 1993.

Gore, A., Jr. "The New Job of the Federal Executive,"Public Administration Review, 54 (July/August ): 317-321, 1994.

Giudice, M. & Goodman, S. "Where the Front and Backend Meet: Collaboration between Designers and Engineers," presentation at the CNET Builder.com Live Conference, December 7-11, 1999.

Harmon, P., Rosen, M. & Guttman, M., Developing E-Business Systems & Architectures: A Manager's Guide, San Francisco: Morgan Kaufmann Publishers, 2001.

Howell, G., Building Hypermedia Applications: A Software Development Guide, McGraw-Hill, Inc, 1992.

Hsian, J. "Web Conferencing of the Local Governments and Democratice Administration," (in Chinese), presented at the "Democratic Administration and Reinventing Government" Colloquium, Shih-Hsin University, Taipei, May 5th, 1999.

Instone, K., " Site Usability Evaluation," Web Review, Available: http://www.webreview.com/1997/10_10/strategists/10_10_97_1.shtml (Oct. 10, 1997).

Jones, S. (ed.), Doing Internet Research: Critical Issues and Methods for Examining the Net, Thousand Oaks: Sage Publications, 1999.

Kilian, C. Writing for the WEB, WA: Self-Counsel Press, 1999.

Lemay, L. Teach Yourself Web Publishing with HTML 4.0 in 14 Days, Professional Reference Edition, IN: Sams.net Publishing, 1998.

Lynch, P. & Horton, S. Web Style Guide: Basic Design Principles for Creating WWW Sites, London: Yale University Press, 1999.

Menzel, D. C. "WWW Ethics.Gov: Issues and Challenges Facing Public Managers, " Public Administration Review, 58 (5): 445-52, 1998.

Mitra, A. & Cohen, E. "Analyzing the Web Directions and Challenges," In Doing Internet Research , Steve Jones, ed., Sage Publications,1999.

Mok, C. Designing Business: Multiple Media, Multiple Disciplines,San Jose, CA: Adobe Press, 1996.

Morris, M. & Ogan, C. "The Internet as a Mass Medium, " Journal of Communication, 46(1): 39-51, 1996.

Nielsen, J.Designing Web Usability: the Practice of Simplicity.Indianapolis, IN: New Riders Publishing, 2000.

Nielsen, J. "The Top Ten New Mistakes of Web Design."Jakob Nielsen's Alertbox. Available: http://www.useit.com/alertbox/ 990530.html (May 30, 1999)

Nielsen, J.Designing Websites With Authority: Secrets of an Information Architect,Indianapolis, IN: New Riders, 1998.

Nielsen, J. "The Need for Speed."Jakob Nielsen's Alertbox. Available:http://www.useit.com/alertbox/9703a.html (March 1,1997).

Nielsen, J. "Top Ten Mistakes in Web Design," Jakob Nielsen's Alertbox. Available: http://www.useit.com/alertbox/9605.html (May 30, 1996).

Nunn, S. & Rubleske, J."“Webbed” Cities and Development of the National Information Highway: The Creation of World Wide Web Sites by City Governments," Available: http://www.spea.iupui.edu/cupe/news/rpt-cvr.htm (December 15, 1998)

Osborne, D. & Gaebler, T. Reinventing Government: How the Entrepreneurial Spirit is Transforming the Public Sector from Schoolhouse State House, City Hall to Pentagon. Reading, MA: Addison-Wesley, 1992.

Parker, R. Guide to Web Content and Design, N.Y.: MIS Press, 1997.

Rosenfeld, L. "15 Site Blueprints: The Key to Good Web Hygiene."Available: http:// www.webreview.com/95/09/15/design/arch/sept15/index.html (September 15, 1995).

Rosenfeld, L. & Morville, P. Information Architecture for the World Wide Web, California: Oreilly, 1998.

Rubin, J. Handbook of Usability Testing: How to Plan, Design, and Conduct Effective Tests,New York: Wiley, 1994.

Scavo, Carmine & Shi, Yuhang , "World Wide Web Site Design and Use in Public Management. " In Information Technology and Computer Applications in Public Administration :Issues & Trends, G. David Garson, ed., Hersey: IDEA Group Publishing., 1999.

Speigel, R. "Mistakes Could Cost E-tailers $6 Billion This Holiday Season."E-Commerce Times.Available: http:// www.ecommercetimes.com/ news/ articles/991008-2.shtml (October 8, 1999).

Wei, C. L., Reinventing Government Movement, Taipei: Morning Star Publications, 1999.

Wiig, K. M. Knowledge Management Foundations: Thinking about Thinking, How People and Organizations Create, Represent, and Use Knowledge, Arlington, Tex: Schema Press, 1993.

Yin, R. K. and Heald, K. A. "Using the Case Survey Method to Analyze Policy Studies," Administrative Science Quarterly,

（本文刊登於Government Information Quarterly 18（2001）357-373.）