WEBSITE NAVIGATION: AN EXPLORATORY STUDY

OF THREE NAVIGATION TOOLS FOR SIMPLE WEB TASKS

Naureen Nizam, Carolyn Watters and Anatoliy Gruzd

Faculty of Computer Science, Dalhousie University, Halifax, NS, Canada

Keywords: Website Navigation, Search, Menu, Tag Cloud, User Experience, User Study, Mixed Methodologies.

Abstract: The purpose of this study is to compare how users perceive and use three navigation tools (search, menus

and tag clouds) on websites to find information. Furthermore, the study compares two variations of tag

clouds, one where the tags are created by a single author and the other where the tags are created by a

community of authors. Using mixed methodologies, data was captured and analyzed for effectiveness,

efficiency and enjoyment. The results show that overall search and tag clouds are more effective and

preferred among users. Tag clouds, where the tags are created by a single author are more efficient than

those created by a community of authors. The findings from this study provide recommendations for future

research in improving navigation on websites.

1 INTRODUCTION

Today there are estimated to be over 600 million

websites on the Web (Netcraft, 2012) with

approximately 266 million users in North American

alone (Pingdom, 2010). Navigation on the Web can

be broken down into two main components:

navigating to the website and navigating to a

specific content on the website (Lazar, Bessiere,

Ceaparu, Robinso and Shneiderman, 2003). Search

engines such as Google, Yahoo and Bing, help users

navigate to websites. Users also find websites by

following links from other web pages or directly

using URLs. Users visit websites for a variety of

different tasks; from quickly finding a restaurant to

more in-depth tasks in planning. Regardless of the

reason for visiting a particular website, users need to

be able to navigate the site even if they are just

browsing serendipitously. Even though websites

offer different navigation tools to aid users in

finding information such as search, menus, tag

clouds, and breadcrumbs, navigation within a

website is one of the main causes of user frustration

on the Web (Lazar et al., 2003).

The effectiveness of search, as a navigation tool

on websites, depends on the user’s ability to form an

appropriate query and search engines effectiveness

to retrieve relevant results. Other tools are based on

set of links that are either created dynamically or

pre-determined. These links are often grouped and

placed together on a web page usually in a form of a

menu or a tag cloud. When a user clicks on a link,

they are directed to another page or file on the web.

These tools can be problematic as information may

be located deep in the website or difficult to find.

Link based features depend on the quality of

categorization and the fit between the categories and

the user’s task and level of understanding. To help

users better navigate a website, it is important to

understand how users perceive and use current

navigation tools. This understanding will identify

ways to improve website navigation.

2 BACKGROUND

While there has been research on navigation, many

of these studies revolve around navigating to sites on

the Web rather than navigating within a website

itself. Research that has examined website

navigation has demonstrated that navigation on sites

can be improved by providing multiple navigation

techniques or paths to the same content (Tullis,

Connor, LeDoux, Chadwick-Dias, True and Catani,

2005); however, less research has been conducted on

how users perceive and use these navigation tools on

websites. In this study, we want to understand user

behavior and perception of three navigation tools on

websites: search, menus, and tag clouds, shown in

Figure 1.

413

Nizam N., Watters C. and Gruzd A..

WEBSITE NAVIGATION: AN EXPLORATORY STUDY OF THREE NAVIGATION TOOLS FOR SIMPLE WEB TASKS.

DOI: 10.5220/0003894704130417

In Proceedings of the 8th International Conference on Web Information Systems and Technologies (WEBIST-2012), pages 413-417

ISBN: 978-989-8565-08-2

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

Tag Cloud

Figure 1: Three navigation tools on many eyes.

2.1 Search

Search allows the user to enter a query into a text

box and retrieve results based on their query. The

search text box is typically accessible from all pages

of the website. Problems associated with search

include: ability of users to form an appropriate

query, effectiveness of the search engine used on the

website, and the frequency at which the pages are

crawled/indexed by the search engine.

2.2 Menus

Menus can be found on most websites. They present

categories of links in a list format, where each item

in the list is a hyperlink to another web page. The

categories are usually created by a website

designer/owner. They often appear consistently

across all pages of the website and are typically

displayed either horizontally or vertically or both.

They

are often used in hierarchical drop down

navigation menus, where the first level is presented

in a horizontal fashion and the second level

presented in a vertical fashion, also referred to as

expandable menus. A study, conducted by Zaphiris,

Shneiderman and Norman (2002), compared

expandable and sequential menus, and found that

reducing the depth of hierarchies improves

performance in terms of speed and search efficiency.

Problems associated with menus include:

appropriate categorization and vocabulary used for

labeling the items in the menu must reflect the user’s

task and context.

2.3 Tag Clouds

Tag clouds, are a visual presentation of weighted

keywords, labels or tags that link to content on a

website. Tag clouds were first seen in 2004 on

Flickr. The tags in a tag cloud are typically

presented in alphabetical order, however, different

approaches such as random or weight based

arrangements have been used (Schrammel, Leitner

and Tscheligi, 2009). Tag clouds usually co-exist

with other navigation tools including menus and

search. They are now found on a wide variety of

different types of websites including personal and

commercial websites, blogs, and social information

sharing sites.

The creation of the tags is often a user-generated

process, where users manually label items on web

pages with related keywords. Depending on the type

of the website, tags may get assigned to items by the

website owner/author or by a community of

authors/users. Tag clouds can also be automatically

created by algorithms based on content. Problems

associated with tag clouds include: reliable tagging

of content, the understanding between the user

assigning the tag and the audience using the tag, and

the tag having a number of different meanings to

different people.

Tagging has been studied extensively in the past

few years and is considered one of the successful

discovery tools, where discovery involves the usage

of social tools to find information (Shneiderman,

2011). Improvements for tagging media objects have

been explored by recommending tags based on

personalized and social context (Rae, Sigurbjornsson

and Zwol, 2010). Zubaiga, Korner, and Strohmaier

(2011), examined automated classification of

content, based on social tagging and user behavior.

Most of the previous research on website

navigation has explored navigation tools

individually. However, Sinclair and Cardew-Hall

(2008) studied the usefulness of tag clouds versus

search for different types of tasks and found that tag

clouds are useful for non-specific information

seeking tasks and search is useful for specific

information seeking tasks. Their research did not

include other navigation tools, such as menus, nor

how users perceive and use navigation tools.

3 OBJECTIVES AND RESEARCH

QUESTIONS

The first objective of this study was to examine how

users find information on websites using three

navigation tools (search, menu and tag cloud) for

simple information seeking tasks, as measured by

effectiveness (task completion) and enjoyment (user

WEBIST2012-8thInternationalConferenceonWebInformationSystemsandTechnologies

414

preference). The second objective of this study was

to compare differences in efficiency (time and

number of clicks to complete the task) between the

three navigation tools and then the two types of tag

clouds; one where the tags are created by a single

author (SA) and the other where the tags are created

by a community of authors (CA). The research

questions are:

1. How do users use three navigation tools

(search, menu and tag cloud) when presented

with all three for simple information seeking

tasks?

2. Which of these three navigation tools are more

effective in finding information for simple

information seeking tasks?

3. Which of these three navigation tools do users

prefer for simple information seeking tasks?

4. Which of the two types of tag clouds (SA or

CA) is more efficient for simple information

seeking tasks?

4 METHODOLOGY

This study was conducted using mixed

methodologies, where both qualitative and

quantitative data was collected and analyzed.

Qualitative data was captured using questionnaires

and interviews, whereas quantitative data (time and

number of clicks) was collected using the Morae

software.

4.1 Participants and Study Design

Fourteen Computer Science students (thirteen males

and one female) participated in this study. The study

was conducted in a lab environment where

participants used Firefox 3.6 to accomplish the tasks.

Participants filled out a background

questionnaire consisting of demographic questions

and previous experience of using navigation tools on

websites, followed by a training session on the three

navigation tools in random order. Participants were

then asked to conduct four tasks. The websites along

with their tasks were presented in random order. For

tasks 1 and 3, participants could use any (or all) of

the three navigation tools. For tasks 2 and 4,

participants were allowed only the tag cloud. After

each task, participants completed a post-task

questionnaire and after all four tasks, they completed

a post-study questionnaire. The study concluded

with an interview session, where participants were

asked a series of questions to elaborate on their

experience of using the three navigation tools.

4.2 Websites and Tasks

The study required four websites, consisting of the

three navigation tools (search, menu and tag cloud)

and the tag cloud of type SA or CA. A focus group

was used to select the four websites and tasks. The

websites and tasks are illustrated in Table 1.

Table 1: Websites, tasks and the type of tag cloud.

Website Description

Type of

Tag Cloud

Professional

on the Web

Companies/professionals

create their profile and also

create tags for their own

services.

Task: Find two companies

that you would consider

hiring to redesign a website.

URL: http://www.professionalontheweb.com/

Many Eyes

Users create/share

visualizations. Users assign

tags to their own

visualizations.

Task: Find two visualizations

beneficial for your project on

alcohol consumption and its

effects.

URL: www-958.ibm.com/

Web

Designer

Wall

A blog of web design ideas,

trends and tutorials.

Maintained by a single author

who tags the blog entries.

Task: Find two posts on how

to design a website for

mobile devices.

URL: http://webdesignerwall.com/

Technology

Education

Know-How

A blog to help teachers use

technology. Maintained by a

single author who tags the

blog entries.

Task: Find two posts to

assist your friend (4th grade

teacher) with resources to

supplement his teaching.

URL: http://www.techedknow.com/

5 SUMMARY OF FINDINGS

The analysis from both qualitative and quantitative

data suggests that search and tag clouds, especially

where tags were created by a single author, were

effective, efficient and enjoyable. Table 2 provides a

summary of the findings.

Additional findings from the interview session

revealed participants’ likes and dislikes on the three

navigation tools.

Search: Participants preferred the flexibility of

forming a search query based on information need.

Furthermore, they found it easy to use. They disliked

the time it required to enter the query (keywords)

WEBSITENAVIGATION:ANEXPLORATORYSTUDYOFTHREENAVIGATIONTOOLSFORSIMPLEWEB

TASKS

415

Table 2: Summary of findings based on research

questions.

Research

Questions

Search Menu

Tag

Cloud

1. How do users

use these tools?

32% 11% 11%

2,3. Perception on:

Effective:

43% 14% 43%

Easiest:

50% 14.29% 35.71%

Preferred:

42.86% 28.57% 28.57%

Efficient:

28.57% 28.57%

42.86%

4. Which of the two types of Tag Cloud (SA/CA) is more

efficient?

Time to

Complete

(ANOVA):

SA Tag Clouds are efficient than CA Tag

Clouds at significance level 0.075

Number of

Clicks

(ANOVA):

SA Tag Clouds required fewer clicks than

CA Tag Clouds at significance level 0.002

(statistically significant)

and the retrieval of sometimes non-relevant results.

Menus: Participants found menus helpful due to

their structured format. They were familiar with the

tool and thought menus were easily accessible, due

to its prominent location on most websites. They

disliked using menus when they are poorly labeled

or when they are not exhaustive.

Tag Clouds: Participants thought tag clouds were

efficient when the appropriate tags (pertaining to

their goal) are present in the tag cloud. They also

thought that tag clouds are well suited for certain

types of websites, such as news and blog websites.

They also indicated that tag clouds are not effective

on their own and should co-exist with other

navigation tools on websites. None of the

participants understood what the tag cloud

represented; majority stated that it might be the most

searched keyword or the most popular content.

Participants disliked similarity between the tags and

tags that are too small to read.

6 RECOMMENDATIONS AND

FUTURE WORK

Search was perceived to be the easiest and preferred,

whereas tag cloud was perceived to be efficient. This

suggests an integration of the two (search and tag

clouds) to allow users the ability to search within a

tag. For example, a user selects a tag and a search

text box is displayed next to the tag to enter a search

keyword(s). Moreover, it might be useful to allow

users to select multiple tags within a tag cloud to

form a search query.

To eliminate the issue of similarity between the

tags in a tag cloud, selecting a tag from a list of

recommended tags during the tagging stage should

be considered. In addition, it would be interesting to

explore the effects of having the community

participate in the selection of the tags. This approach

to tagging could be presented in a tag cloud format

and the use of appropriate algorithms.

Menus were not ranked as the most useful

navigation tool in any category and this opens up

possibilities for more dynamic construction of

menus, where menus would be updated in real-time

based on the website’s activity (Google Analytics)

or community input (Twitter and Facebook).

7 CONCLUSIONS

Navigation tools within websites should be easily

accessible and lead to relevant information quickly.

The findings from this study, both qualitative and

quantitative, reveal that users perceive tag clouds as

useful as search in finding information. Tag clouds

where the tags are created by a single author are

more efficient (in terms of number of clicks) than

tags created by a community of authors. The study

provides sufficient evidence to investigate and

improve navigation on websites including tag

clouds, especially where websites rely on

community driven content and tagging.

REFERENCES

Lazar, J., Bessiere, K., Ceaparu, I., Robinson, J. and

Shneiderman, B. (2003). Help! I’m lost: User

frustration in web navigation. IT&Society: A Web

Journal Studying How Technology Affects Society,

1(3), 18-26.

Netcraft. February 2012 Web Server Survey. Retrieved

February 9, 2012, from http://news.netcraft.com/

archives/category/web-server-survey/

Pingdom. Internet 2010 in numbers. Retrieved February 9,

2012, from http://royal.pingdom.com/2011/01/12/inter

net-2010-in-numbers/

Rae, A., Sigurbjornsson, B., and Zwol, R. (2010).

Improving tag recommendation using social networks.

Proceedings of the RIAO’10 Adaptivity,

Personalization and Fusion of Heterogeneous

Information. Paris, France.

Schrammel, J., Leitner, M., and Tscheligi, M. (2009).

Semantically structured tag clouds: An empirical

evaluation of clustered presentation approaches.

Proceedings of the 27th international conference on

Human Factors in Computing Systems, 2037-2040.

ACM Press.

Shneiderman, B. (2011). Social discovery in an

information abundant world: Designing to create

WEBIST2012-8thInternationalConferenceonWebInformationSystemsandTechnologies

416

capacity and seek solutions. Information Services &

Use, 31(1-2), 3-13.

Sinclair, J. and Cardew-Hall, M. (2008). The folksonomy

tag cloud: When is it useful?. Journal of Information

Science, 34(1), 15-29.

Tullis, T., Connor, E., LeDoux, L., Chadwick-Dias, A.,

True, M., and Catani, M. (2005). A study of website

navigation methods. Paper presented at Usability

Professionals Association (UPA) 2005 Conference,

Montreal.

Zaphiris, P., Shneiderman, B. and Norman, K. (2002).

Expandable indexes vs. sequential menus for

searching hierarchies on the World Wide Web.

Behaviour and Information Technology, 21(3), 201-

207.

Zubaiga, A., Korner, C., and Strohmaier, M. (2011). Tags

vs shelves: From social tagging to social classification.

Proceedings of the Hypertext 2011, the 22nd ACM

Conference on Hypertext and Hypermedia, 93–102.

June 6-9, 2011, Eindhoven, Netherlands. ACM Press.

WEBSITENAVIGATION:ANEXPLORATORYSTUDYOFTHREENAVIGATIONTOOLSFORSIMPLEWEB

TASKS

417