THE EFFECTS OF THE AJAX TECHNOLOGY ON WEB

APPLICATION USABILITY

Jonas Kluge

Smile - Motoristes Internet, Paris, France

Frank Kargl, Michael Weber

Institute of Media Informatics, Ulm University, Germany

Keywords:

Ajax, Usability, WWW, Web Application, Efﬁciency, Satisfaction, Survey.

Abstract:

The Ajax technology is said to make the Internet faster, more interactive and user-friendly. It is spreading

rapidly throughout the web. By allowing scripts being executed inside the user’s browser to communicate

with a remote server, it enables new forms of interaction for web-pages and -applications. In this paper we

analyze Ajax usability, taking into account both advantages and imposed problems. Subsequently we present

a user study with two typical scenarios in web-applications and compare Ajax-enabled and non-Ajax-enabled

versions. This allows us to verify and quantify the assumed effects. As a result, we show that Ajax signiﬁcantly

enhances both user satisfaction and efﬁciency of use, at least in some scenarios.

1 AJAX INTRODUCTION

The term Ajax is an acronym for “Asynchronous

JavaScript and XML” (Garrett, 2005) and describes

a way to realize a HTTP communication from within

a JavaScript program embedded into a web page.

The retrieval of information or the usage of an

application in traditional websites consists of load-

ing web pages page-by-page by clicking on links or

submitting forms. Once a page has been loaded, the

content remains mostly unchanged until a user action

triggers loading of a completely new page.

In contrast to this, Ajax sites can dynamically

load new data from and transmit data to the originat-

ing web server while the page stays in display. The

Javascript application running inside the browser can

update the content of the web page with no need to

discard the whole page at every data transmission.

Communication happens asynchronously in the back-

ground, the web application stays fully functional in

the meantime. Figure 1 shows the control-ﬂow of an

Ajax site compared to a regular web application.

The communication with a webserver in an Ajax

application is controlled by the so-called XML-

HttpRequest object, which offers either synchronous

or asynchronous retrieval of data. In the latter case,

the application is notiﬁed via callbacks, when the re-

trieved data is available. The XMLHttpRequest ob-

ject is an API being accessible from major scripting

languages, like e.g. JavaScript. Its runtime environ-

ment is the Ajax engine being embedded into a stan-

dard web browser.

Using the available mechanisms, highly inter-

active web applications can be realized, like e.g.

Google Maps (http://maps.google.com) where users

can seamlessly navigate through maps or Gmail

(http://www.gmail.com/) where the inbox can be up-

dated or emails moved between folders, all without

loading a new page.

So Ajax-enabled web applications seem to offer

huge beneﬁts in usability compared to traditional web

applications. But studies trying to analyze this are

very rare. So our goal is to provide some signiﬁcant

results that substantiate certain aspects of Ajax usabil-

ity.

2 USABILITY ASPECTS

When used in web pages, Ajax enables new interac-

tion modes with web pages, which imply effects on

the usability of these. In this section we will analyze

the improvements that an adequate use of Ajax can

289

Kluge J., Kargl F. and Weber M. (2007).

THE EFFECTS OF THE AJAX TECHNOLOGY ON WEB APPLICATION USABILITY.

In Proceedings of the Third International Conference on Web Information Systems and Technologies - Web Interfaces and Applications, pages 289-294

DOI: 10.5220/0001286102890294

 SciTePress



Time

Browser

HTML page

Server

Time

Interface

HTML page

Interface

HTML page

Interface

Classic Web Application

Browser

HTML page

Ajax Engine 

Server

Interface

Display Input Display Input Display

Req.‐



Response 

Req.



Response 

Req.



Response

Req.



Response

Req.



Response

Req.



Response 

Req.



Response

Ajax application

Figure 1: Comparison of classic web applications and ajax applications.

mean for websites; in the following section we will

discuss the drawbacks of its usage.

2.1 Application Speed

Web sites can use Ajax to load content requested by

the user. Given a web page is already present at the

users browser, subsequent loads can be truncated to

the changes. This way the server can omit sending the

code used for layout and unchanged parts of the page

(like menus) and just transfer the new content. This

can reduce the loading time of a page signiﬁcantly.

The effect is most noticeable when only a rela-

tively small part of a big page changes or is addi-

tionally loaded. Also when a form is submitted as

a part of a page and the rest of the page is supposed

to remain in display (e.g. the later examined message

board with an additional message being added), the

transfer via Ajax can save transmission time.

2.2 User Interface Smoothness and

Interactivity

As the user can work within a page and the displayed

page doesn’t have to be discarded at each request of

additional content, the interaction with the web page

can be designed substantially more ﬂuid than before.

The interruption of the user’s work process and the

required adaption to a new workspace which happen

when the page is changed disturb the workﬂow of the

user. When content is sent or loaded via Ajax, the user

interface remains continuously visible and unchanged

apart of the new content elements.

A good user interface interacts with the user and

reacts instantaneously to his actions. Lowry et al.

compared in a study variants of web applications with

different grades of interactivity and found out that the

satisfaction of the user increases with higher interac-

tivity. (Lowry et al., 2006)

Regarding the interactivity, web applications were

always in an inferior position compared to desktop

software. Classic web applications can only directly

react to user actions as far as the required informa-

tion had been delivered to the browser with the page.

In any other case a new page has to be loaded. This

limits potential reactions to predictable user input. In

case the reaction of the application is based on a big

amount of data it is neither useful nor wanted to trans-

mit all data to the user. Using Ajax the application

can access server side data from inside the page and

react to user input dynamically. Thus only these parts

of data which are relevant to handle the current user

input have to be loaded.

WEBIST 2007 - International Conference on Web Information Systems and Technologies

290

2.3 Data Transfer Transparency

We have already mentioned that Ajax can reduce the

actual time the user has to wait for new content to

appear. Often the subjective impression of waiting

depends not so much on the actual time it takes to

load and display the content but on the way the user

interface presents it. Culwin and Faulkner (Culwin

and Faulkner, 2001) found out that users preferred

an interface keeping them informed regarding the to-

tal loading progress over an interface with a regular

browser loading indicator.

With Ajax, the technical means now exist for web

applications to provide custom loading indicators to

the user. By being able to control the user interface

without interruption, the application can keep the user

informed about when loading occurs, what is being

loaded and what the current progress is. In contrast

to this, regular web browser UIs only display a load-

ing animation and a rough estimation of the loading

progress.

2.4 New Interactions

Ajax can not only replace regular loading of content

and speed up applications by this, but also enables

new interactions with a web page. For example an in-

put ﬁeld which offers completion suggestions for the

typed characters can hardly be realized without Ajax.

Neither a reload of the whole page after each typed

character nor preloading all possibly needed data (like

the train stops in the later mentioned travel planner)

together with the web page would be very useful.

3 PROBLEMS

As we have seen, the use of Ajax can improve website

usability in several modes. Unfortunately, by chang-

ing the way the web applications work, problems arise

as well. In this section we discuss the drawbacks of

Ajax in the ﬁelds of accessibility and the disturbance

of web browser navigation tools.

3.1 Accessibility

As Ajax applications rely completely on JavaScript,

they are not accessible to users of browsers, which do

not support this language. Text browsers such as Lynx

and browsers for mobile devices but also screenread-

ers for visually impaired are part of this group.

As the different web browsers vary in extent and

implementation of Javascript support, the risk is high

that Ajax applications do not work as expected with

some browser versions. To reach a broad compatibil-

ity, strong efforts are necessary.

Possible solutions to the accessibility problem in-

clude the delivery of normal HTML pages, which are

subsequently enriched with Ajax functionality. To

realize this, a non-Ajax-based version of the page

is loaded and subsequently modiﬁed by a JavaScript

programme, if Ajax support has been detected.

3.2 Browser Navigation and Bookmarks

In normal web applications, users navigate between

pages and can use the browser’s “back” and “forward”

buttons to return to recently visited items. Typically

pages can also be bookmarked and re-visited later.

As with Ajax the change of page content is con-

trolled by scripts, the browser cannot simply go

“back” any more. Clicking the navigation buttons

might lead to undesirable consequences.

Also, when content is loaded via Ajax, the URI

of the page does not change, which renders the book-

mark function of web browses useless. Accessing a

bookmark of an Ajax page thus loads the initial screen

only.

Several workarounds for these problems are

known. A promising approach to the navigation prob-

lem is the implementation of custom functions for the

handling of the “back” and “forward” buttons, which

is supported by the current releases of Mozilla Fire-

fox and Microsoft Internet Explorer browsers. The

bookmark problem can be solved by coding the cur-

rent page-state into the in-page link anchor marker

“#” either by providing a link to the current state or

by automatically jumping to a new virtual anchor on

each page update.

4 EMPIRICAL SURVEY

To support our theoretical ﬁndings from the last sec-

tions, we conducted a user testing in which we mea-

sured the differences in selected usability aspects be-

tween Ajax and non-Ajax-versions of two web appli-

cations. This chapter describes the conﬁguration of

the study and its results.

4.1 Subjects and Measures

We decided to examine two different usages of Ajax:

On the one hand we choose a form ﬁeld auto-

complete widget as a very typical example for appli-

cations that become possible when using Ajax. The

auto-complete widget is used in a public transit trip

planner, where the ﬁelds for the start and destination

THE EFFECTS OF THE AJAX TECHNOLOGY ON WEB APPLICATION USABILITY

291

station suggest completions for the typed characters,

based on a database of all train stations in Germany.

In our second example, Ajax is used to send the

content of a form to the server in order to save it to a

database while the user can continue to use the appli-

cation. This is implemented in a discussion forum ap-

plication where the user can add a comment which is

submitted to the server via Ajax and directly inserted

into the tree-like message board. In the meantime,

the user can continue reading or posting to the forum

without interruption.

While in the ﬁrst example the expected advan-

tage consists in the support of the user through the

auto-complete suggestions, the main issues in the sec-

ond example are the reduced transfer time and the in-

creased smoothness of the user interface.

For the study we implemented both web applica-

tions in a version with and a version without Ajax

usage. As measures we choose two sub-aspects of

usability: satisfaction and efﬁciency (Nielsen, 1993).

We select satisfaction, as it is a subjective impression

and summarizes the whole experience of the user with

the application. In contrast, efﬁciency is chosen as an

objectively measurable criterion.

Satisfaction was measured with a 12 item ques-

tionnaire based on the IBM Usability Satisfaction

Questionnaire (Lewis, 1995) with statements like

“Overall, I am satisﬁed with how easy it is to use

this system”Each item has to be rated on a scale with

ﬁve possible gradings between “strongly disagree”

and “strongly agree”. Efﬁciency was quantiﬁed by

the time the users needed for the completion of given,

characteristical tasks.

4.2 Hypotheses

According to the chosen subjects and measurements,

four hypotheses are tested:

H1: The user satisfaction with the trip planner with

auto-completion of input ﬁelds is higher than without

auto-completion.

H2: The efﬁciency of using the trip planner with auto-

completion of input ﬁelds is higher than without auto-

completion.

H3: The user satisfaction with the message board with

transmission of new comments via Ajax is higher than

with regular comment transmission.

H4: The efﬁciency of using the message board with

transmission of new comments via Ajax is higher than

with regular comment transmission.

4.3 Methodology

The study was conducted via Internet. The partici-

pants were asked to complete a set of tasks for each

application. Each user worked with both the Ajax and

the non-Ajax version of the two applications, so they

had to complete four sequences altogether.

The order of the applications and the Ajax and

non-Ajax version was randomized to eliminate suc-

cession effects. Two different variants of the tasks

were created for the two versions of the applications,

the assignment of the task variants to the application

versions was randomized, too.

After each version the participant was requested to

ﬁll out a satisfaction questionnaire. At the end of the

survey statistical data (age, gender, experience with

computers and the Internet) was gathered and a text

ﬁeld for transmitting feedback was offered.

4.4 Results

The survey has been completed by 123 participants

successfully. Participants were between 14 and 68

years old with an elevated concentration between 20

and 30. There was a slight dominance of males with

a rather high experience with computers and the in-

ternet. Figure 2 shows the age and gender distribu-

tion. A small number of participants (exceeding the

123) experienced technical problems, mainly browser

compatibility issues, and have not been included in

the analysis.

The results show different extents of improvement

on satisfaction and efﬁciency in the two applications.

For the statistical validation of the results we applied

an analysis of variance for a repeated measures de-

sign. Using this method, three of the four hypotheses

can be conﬁrmed while the fourth cannot be validated

statistically.

Figure 3 gives an overview over the results. Figure

4 shows a histogramm of the time required to com-

plete the given tasks, which is taken as the inverse of

efﬁciency.

The satisfaction with the trip planner improves

from 3.7 to 4.37 (on a scale from 1 to 5) when in-

troducing the Ajax-enable auto-complete widget. The

corresponding hypothesis H1 can be proven with high

signiﬁcance (p < 0.01).

The time needed to complete the given tasks with

the trip planner decreases in average from 92.5 s to

71.5 s with Ajax auto-completion which shows a clear

gain in efﬁciency. Hypothesis H2 can thus be corrob-

orated also with high signiﬁcance (p < 0.01).

In the message forum application, the satisfaction

of the users improve from 3.84 to 4.22 when the com-

WEBIST 2007 - International Conference on Web Information Systems and Technologies

292

AgeDistribution

GenderDistribution

female;

‐

male;81

Figure 2: Age and gender distribution of the participants.

92,5

RequiredTime

withAjax withoutAjax

4,37

4,22

384

Satisfaction

withAjax withoutAjax

71,5

63,8

70,0

3,70

tripplanner messageboard

Figure 3: Results of the study: Measured Satisfaction for the Ajax and the non Ajax version of the two evaluated applications

(higher value = better) and time in seconds needed to complete the given tasks (lower value = better).

ments are transmitted via Ajax. Hypothesis H3 can

also be conﬁrmed with high signiﬁcance (p < 0.01).

The efﬁciency also increases when using of Ajax,

reducing the average time needed for the tasks from

70 s to 63.8 s. However, the statistical test cannot con-

ﬁrm the corresponding hypothesis H4 (p = 0.067).

Summed up the employment of Ajax leads to a

satisfaction gain in both cases, whereas the efﬁciency

beneﬁts can only be conﬁrmed in one case. These

results will be discussed and interpreted next.

4.5 Interpretation of Results

The results show that the introduction of Ajax leads

to improvements of different extents in terms of satis-

faction and efﬁciency of use of web applications.

The auto-completion function has clearly proven

its advantages in terms of satisfaction as well as efﬁ-

ciency. The so improved trip planner received many

positive comments from the users, which requested to

have this function available in real trip planner appli-

cations.

In the second application, which uses Ajax to

transmit the content of a form, the advantages are not

as big as with the auto-completion. Regarding efﬁ-

ciency there was no signiﬁcant gain at all. A possi-

ble explanation is the fact that the internet connection

speeds in the tests were so fast, that the time advan-

tage reached through the transmission via Ajax was

hardly noticeable for the majority of the users.

Another reason for the failure to signiﬁcantly

prove the efﬁciency gain is the high variance in time

needed by the participants to complete the tasks.

Though being instructed to quickly scan the forum

and write just anything, many participants read the

comments thoroughly and tried hard to write mean-

ingful contributions. A similar trial should be pos-

sibly conducted using a different application sce-

nario. Anyhow the Ajax-based version did not per-

form worse than the non-Ajax one.

5 CONCLUSION

In this paper we presented many advantages when

using Ajax in web applications but also considered

some drawbacks. The main advantages in the ﬁeld

of usability affect user satisfaction and response time

THE EFFECTS OF THE AJAX TECHNOLOGY ON WEB APPLICATION USABILITY

293

Required time for the trip planner

with Ajax

without Ajax

Required time for the message board

with Ajax

without Ajax

Figure 4: Histogramm of the time required to complete the given tasks with the two applications, comparing the Ajax and non

Ajax version each. A lower value indicates a higher efﬁciency.

which inﬂuence the efﬁciency with which the applica-

tion can be used. Besides this, new functions enhanc-

ing user comfort can be created. Some new kinds of

applications are only possible through the use of Ajax.

On the other hand problems may arise because

the browser loses some control over the HTTP trafﬁc

and its functions for navigation may not work as ex-

pected. However, with some effort developers can re-

alize working browser navigation functions for most

scenarios.

A more fundamental problem comes up when a

user cannot use an Ajax application at all because her

browser is incompatible with Javascript and/or Ajax.

This is also an aspect of usability and maybe the most

severe usability problem: The non-usability. When

designing an application developers have to take into

consideration the target group: will it be able to use

the application or is partial non-accessibility viable.

Unquestionably Ajax makes sense when creating

comfort functions like the auto-completion described

previously. They are attached to existing applica-

tions and do not replace any regular HTTP transfer

but only add visual support and convenience. Thus,

they can easily be omitted if the browser does not sup-

port Ajax. Then, Ajax can be used without any of the

disadvantages being mentioned.

For the group of applications that cannot be real-

ized without Ajax, the situation is different. They may

not be accessible for some groups of Internet users,

but compared to the former non-existence of this ap-

plication class, it is still an improvement.

5.1 Future Work

While we looked at the beneﬁts of Ajax not only the-

oretically but also empirically in our study, we ana-

lyzed the drawbacks only in theory. In a future work

it could be attempted to formalize and quantify the

mentioned problems of the Ajax technology. Other

future research could deal with a more comprehensive

evaluation of other possible applications and include

other usability aspects.

Ajax spread rapidly the last years and will con-

tinue to do so. Our paper showed that Ajax is no

magic bullet as it can bring up new problems. How-

ever we could also show that the employment of Ajax

brings a noticeable gain of efﬁciency and satisfaction

in certain ﬁelds. Wisely used, Ajax can clearly im-

prove the usability of web applications that way.

REFERENCES

Culwin, F. and Faulkner, X. (2001). Brewsing the web:

Delay, determination and satisfaction. In HICSS ’01:

Proceedings of the 34th Annual Hawaii International

Conference on System Sciences, page 5018, Washing-

ton, DC, USA. IEEE Computer Society.

Garrett, J. J. (2005). Ajax: A new

approach to web applications.

http://adaptivepath.com/publications/essays/archives/

000385.php.

Lewis, J. R. (1995). Ibm computer usability satisfaction

questionnaires: psychometric evaluation and instruc-

tions for use. Int. J. Hum.-Comput. Interact., 7(1):57–

78.

Lowry, P., Madariaga, S., Mofﬁt, K., Moody, G., Spauld-

ing, T., and Wells, T. (2006). A theoretical model

and empirical results linking website interactivity and

usability satisfaction. HICSS ’06. Proceedings of the

39th Annual Hawaii International Conference on Sys-

tem Sciences, 6:123a.

Nielsen, J. (1993). Usability Engineering. Academic Press.

WEBIST 2007 - International Conference on Web Information Systems and Technologies

294