Web-based Search: How Do Animated User Interface Elements Affect

Autistic and Non-autistic Users?

Alexandra L. Uitdenbogerd, Maria Spichkova and Mona Alzahrani

School of Computing Technologies, RMIT University, Melbourne, Australia

Keywords:

Search, Information Retrieval, Autistic Users, Animation, Usability, HCI.

Abstract:

Many websites and other user interfaces include animated elements, particularly for advertisements. However,

these can have a negative impact on users, with some cohorts, such as autistic users, being more affected. In

our mixed methods study on the effect of irrelevant animations on usability we observed the effect on search

activities. For those greatly impacted by on-screen animation the effect was not always to slow down a task,

but search terms were entered hastily to avoid more exposure, with shorter queries on average and a greater

tendency to copy and paste during query formulation. Autistic users found the task more mentally demanding,

and were more distracted or annoyed by the animations.

1 INTRODUCTION

Animations form part of many user interfaces, partic-

ularly on websites. However, they are known to inter-

fere with a user’s ability to complete their tasks (Hong

et al., 2007). There has been much research on

the impact of animated banner advertisements on the

user experience, with particular attention to potential

sales (see for example Lohtia et al., 2003; Rau et al.,

2006; Yoo et al., 2004; Zorn et al., 2012). There are

fewer studies that examine how well users achieve

their goals on web-sites in the presence of on-screen

animations. There is some evidence that users’ ability

to complete tasks is more impacted for browsing than

search, with the former considered a cognitively more

demanding task (Cheung et al., 2017; Pagendarm and

Schaumburg, 2001; Wang et al., 2014; Hong et al.,

2007).

Some users may be more impacted than others

by having animated elements in the user interface.

Indeed, even for static pages, there is evidence that

autistic users have more difﬁculty focusing on the

relevant screen elements in order to achieve their

goals (Eraslan et al., 2019).

According to Vos et al. (2016), approximately 1%

of the overall population might be on on the autism

spectrum. Individuals diagnosed with an autism spec-

trum condition often have hypersensitivity or hy-

posensitivity to sensory inputs, in addition to atypical

attention tendencies (Schaaf and Miller, 2005). There

is some evidence from eye-tracking experiments that

autistic users are more likely to be distracted by ir-

relevant elements on a user interface (Eraslan et al.,

2019). However, despite moving content being stated

as a barrier to accessibility

, current web development

guidelines addressing accessibility for those with cog-

nitive differences appear to be largely based on anec-

dotal evidence (Seeman and Cooper, 2015).

Contributions: In this paper we report on part of a

study that examined the impact of irrelevant on-screen

animation on two cohorts of users’ ability to complete

a range of tasks (Alzahrani et al., 2021, 2022). Here

we report the results and additional analysis related to

the search task from the study to answer the following

research question:

RQ: How does the effect of irrelevant animations

within user interfaces vary between autistic and non-

autistic users, while conducting Web search tasks?

The aim of our work was to investigate the inter-

action between distractors and users’ cognitive style

(autistic versus non-autistic), in addition to user per-

formance across various internet-based tasks. To-

gether with wring and reading emails, search is

deemed the most popular online activity (Purcell,

2011) and was therefore an essential task to include

in our study.

Outline: The rest of the paper is organised as follows.

Section 2 introduces the related work. In Section 3,

we discuss the methodology we applied to conduct

https://www.w3.org/WAI/people-use-web/abilities-

barriers/

Uitdenbogerd, A., Spichkova, M. and Alzahrani, M.

Web-based Search: How Do Animated User Interface Elements Affect Autistic and Non-autistic Users?.

DOI: 10.5220/0011074500003176

In Proceedings of the 17th International Conference on Evaluation of Novel Approaches to Software Engineering (ENASE 2022), pages 453-460

ISBN: 978-989-758-568-5; ISSN: 2184-4895

453

the study. In Section 4, we present the results of the

study, which are then analysed and discussed in Sec-

tion 5. The limitations of our study are discussed in

Section 6. Finally, Section 7 summarises the paper.

2 RELATED WORK

In our preliminary work, we presented and dis-

cussed a literature survey on the animation case stud-

ies (Alzahrani et al., 2021). A number of studies in

various disciplines have focused on the inﬂuences of

animations and images on non-autistic users, such as

the work of Hong et al. (2007), whereas there are

only very few studies on autistic users. To our best

knowledge, none of the existing studies has focused

on the analysis of the impact of animated user inter-

face elements on conducting Web-based search tasks

by autistic and non-autistic users.

The guidelines on Web accessibility have been

provided by the World Wide Web Consortium (W3C)

in the form of the Web Content Accessibility Guide-

lines (WCAG) Version 2.1. However, users with cog-

nitive impairments confront challenges that were not

focused on so much in WCAG, as catering to phys-

ically disabled users’ needs was the main emphasis

(Yesilada and Harper, 2019).

According to the Weak Central Coherence theory

(Happ

e and Frith, 2006), the cognitive proﬁle of autis-

tic individuals tends to be biased toward local sensory

information processing, rather than semantic, contex-

tual and global information processing. Despite the

local bias, Eraslan et al. (2019) demonstrated that

autistic users had more holistic (and therefore less fo-

cused) eye-tracking patterns than non-autistic users,

leading to potentially lower success in some focused

tasks. Perhaps this is due to the high overlap between

autism and Attention Deﬁcit/Hyperactivity Disorder

(ADHD) diagnoses (Schaaf and Miller, 2005).

A review of user experience studies with autistic

users was presented in C¸ orlu et al. (2017). It covers

98 studies conducted between 2010 and 2016. The

authors applied both qualitative and quantitative ap-

proaches in their meta-analysis. Their results showed

that the most studied cohort are children (47%) with

only 11.9% of studies involving autistic adults. Most

studies were focused on software to address issues

such as social interaction and communication. Very

few, if any, addressed accessibility issues of the Web

or other general software.

A systematic literature review on the impact of

technology on autistic individuals was presented in

Valencia et al. (2019). The authors reviewed 94 stud-

ies to analyse how the use of technology in educa-

tional contexts helps autistic people develop several

skills, how these approaches consider aspects of user

experience, usability and accessibility, and how game

elements are used to enrich learning environments.

As with C¸ orlu et al. (2017), the articles in this review

were largely about educating autistic children, with

accessibility and usability not adequately addressed.

Similarly, there are many other studies focused on

technology solutions for autistic children (Battocchi

et al., 2010; Millen et al., 2010; Sitdhisanguan et al.,

2012; Gentry et al., 2010).

Results from an anonymous on-line survey on the

user experience of software or technology designed

for autistic people were discussed in Putnam and

Chong (2008) but the majority of respondents were

parents or carers of autistic children rather than autis-

tic users themselves. The emphasis in the survey re-

sults was therefore more on solving social, communi-

cation and educational problems rather than speciﬁc

user interface issues.

In summary, there is evidence that irrelevant ani-

mation impacts users and that autistic users are more

likely to be distracted by static irrelevant elements on

a user interface. However, apart from our study, there

does not appear to be other research into how irrel-

evant animation impacts autistic users during search

tasks.

3 EXPERIMENTS

In order to obtain comprehensive insights into users’

experience, our approach was to use mixed methods,

with quantitative and qualitative data collected in par-

allel. We triangulated based on the different types of

evidence collected in the study.

3.1 Procedure

Before commencing the study, approval was obtained

from the relevant university ethics committee. Par-

ticipants who answered recruitment advertisements

and consented to take part initially completed a pre-

questionnaire about their experiences and any diag-

noses that may impact their participation, such as At-

tention Deﬁcit/Hyperactivity Disorder (ADHD) or vi-

sion impairment.

Our aim was to conduct the experiment with par-

ticipants representing both autistic and non-autistic

populations of users, where our inclusion criteria for

classing a participant as autistic were a diagnosis

of autism spectrum condition (ASC) levels 1 and 2

without intellectual disability, or self-identiﬁcation as

autistic.

ENASE 2022 - 17th International Conference on Evaluation of Novel Approaches to Software Engineering

454

Figure 1: Search Screenshot.

Three severity tiers of ASD (ASD level 3, ASD

level 2 and ASD level 1) are deﬁned in the most re-

cent version of The Diagnostic and Statistical Manual

of Mental Disorders (DSM-5) (American Psychiatric

Association, 2013). As per DSM-5, average to su-

perior intelligence is seen among individuals at ASD

level 1 Autism Spectrum Condition (ASC), although

sensory sensitivity and social contexts may neverthe-

less be challenging for them.

An on-line meeting was then held between the

third author and the participant, during which the par-

ticipant completed a set of six tasks, each with a dif-

ferent distractor. Given that users may become tired

as the experiment progresses, it may be this factor

that causes the ﬁnal tasks to potentially be completed

to a worse degree rather than the animation’s impact.

Consequently, the results or ﬁndings of prior condi-

tions may bias the user’s answers or reactions to the

later conditions. In the case of animated elements, the

learning effect means that participants learn to ignore

these distractors.

Hong et al. (2007) state that when users initially

see animation on a website they may ﬁnd it difﬁ-

cult to ignore, but they will learn to partially block

it and focus on the main task. To avoid this threat to

validity, we organised our experiment using a Latin

Square arrangement (Fisher and Yates, 1938), which

is a common approach for experimental design (Ma-

jrashi, 2016; Burke et al., 2005). The Latin Square de-

sign is a technique for ordering conditions and tasks in

a balanced way across study participants. In this case,

the animation conditions were rotated across tasks, so

that each user completed each task with a different

animation condition and subsequent users completed

them with different task-animation-condition combi-

nations to prior users.

Thus, each of the animation conditions consisted

of a rotating university logo, the size and speed of

which were varied. We report here the results per-

taining to the search task, the screen shot for which is

shown in Figure 1. The order of animation is reversed

after every six participants, as shown in Table 1.

Table 1: Latin Square design (Case 1: Black and White

image, Case 2: Slow rotate, Case 3: Coloured image, Case

4: Fast rotate, Case 5: Nothing and Case 6: Small slow

rotate).

Task 1 Task 2 Task 3 Task 4 Task 5 Task 6

case 1 User 1 User 2 User 3 User 4 User 5 User 6

case 2 User 2 User 3 User 4 User 5 User 6 User 1

case 3 User 3 User 4 User 5 User 6 User 1 User 2

case 4 User 4 User 5 User 6 User 1 User 2 User 3

case 5 User 5 User 6 User 1 User 2 User 3 User 4

case 6 User 6 User 1 User 2 User 3 User 4 User 5

case 6 User 7 User 8 User 9 User 10 User 11 User 12

case 5 User 8 User 9 User 10 User 11 User 12 User 7

case 4 User 9 User 10 User 11 User 12 User 7 User 8

case 3 User 10 User 11 User 12 User 7 User 8 User 9

case 2 User 11 User 12 User 7 User 8 User 9 User 10

case 1 User 12 User 7 User 8 User 9 User 10 User 11

The experiment was conducted online, with users’

screens being observed via online tools, such as

screen sharing via Skype. To measure the animation’s

effect on autistic and non-autistic users, the task com-

pletion rate, user error rate and time taken per task

were recorded. User error rate was deﬁned differently

for each task. For search it was deﬁned as a query

with irrelevant words or a blank query. For the anal-

ysis in this paper, we use measures of query success,

described in Section 3.2.

After each task, the participant answered several

questions related to the task they had just completed.

A further questionnaire was answered after all tasks

had been completed. In addition to the data collected

via the web application and on-line survey forms, the

on-line meeting was screen-recorded for later analy-

sis.

3.2 Data Collection and Analysis

We posted an announcement on social media, with

online autism organisations, as well as websites that

Web-based Search: How Do Animated User Interface Elements Affect Autistic and Non-autistic Users?

455

aim to exchange surveys, to recruit participants for

both the autistic and non-autistic cohorts. Twelve

autistic and twenty six non-autistic participants com-

pleted the study. All participants classed as autistic in

this study stated that they had a diagnosis of autism.

Task durations were captured, as well as partici-

pant responses regarding the perceived effort in com-

pleting the task. For the search task, the search query

was recorded. All queries were later tested on a search

engine to determine whether they were successful by

recording the rank of the ﬁrst relevant document.

Two simple metrics were used to compare the dif-

ferent search conditions: Success at 10 and Mean Re-

ciprocal Rank (MRR). The Success at 10 metric gives

a score of 1 when there is a relevant query in the ﬁrst

10 results and 0 otherwise. Thus, the mean across

queries will be a number between 0 and 1. The metric

was selected to represent whether the user’s query was

successful at all, given that a typical search engine

provides ten results per page and most users will only

look at the ﬁrst results page most of the time (Zhang

and Moffat, 2006).

MRR is a simple measure that uses the rank of the

ﬁrst relevant query result to represent relative success

in the search task. For example, if the ﬁrst relevant

result was in position number 2, the reciprocal rank

is 0.5. As results appear later in the ranked list, the

reciprocal rank approaches zero. The mean for both

measures was calculated across participants’ queries

for the one search task rather than across multiple

search tasks.

We chose the DuckDuckGo search engine for

query testing to minimise any personalisation that

may occur during the search session. Location in-

formation was disabled in the browser during search.

Documents were classed as relevant if they referred to

a bachelor degree on circus arts at a university located

in Melbourne, as per the task presented in Figure 1.

In addition, query lengths are reported for each

condition and cohort, as well as whether a query was

typed or cut and pasted from the task description.

4 RESULTS

Table 2 shows the task duration across all partici-

pants for the search query formulation task and two

browsing tasks. We use here the following notation:

N refers to the number of participants, M the mean

and SD the standard deviation. Unlike the other tasks,

search query formulation under the animation condi-

tion was slightly faster on average. Despite the speed,

all queries completed in the animated condition were

successful in retrieving a relevant result within the

Table 2: Task duration in seconds for both conditions (with

and without animation).

Type of Task Animation M SD N

Menu Options No animation 96.52 33.08 19

Animation 109.36 44.99 19

Total 102.94 39.49 38

Selection No animation 74.10 30.11 19

Animation 115.44 51.41 18

Total 94.21 46.26 37

Search Query No animation 59.73 26.18 19

Formulation Animation 55.68 22.30 19

Total 57.71 24.07 38

Table 3: Mean reciprocal rank.

Animation No Animation

Autistic 0.86 0.70

Non-Autistic 0.88 0.94

Table 4: Mean query lengths in characters.

Animation No Animation

Autistic 30.1 31.0

Non-Autistic 29.8 36.8

ﬁrst ten documents with the DuckDuckGo search en-

gine (as judged by the third author). The only unsuc-

cessful query by this measure (“performance arts”)

was formulated by an autistic user in the unanimated

condition.

Table 3 shows that the mean reciprocal rank

only differed greatly for the autistic unanimated case,

which was the impact of the single unsuccessful query

mentioned above. Excluding the outlier changes the

MRR from 0.7 to 0.88. It must be noted that this is a

summary of only four values when the outlier is ex-

cluded. The vast majority of queries (30) had their

ﬁrst relevant result at rank 1, 6 had a rank of 2 and one

a rank of 4. Interestingly, ﬁve of the six queries with

rank 2 were in the animated condition, which may be

why there is a notable difference between the anima-

tion and non-animation scores for the non-autistic co-

hort. The queries with rank 2 either left out the word

Melbourne (for example, “circus arts degre”) or used

“art” instead of “arts” (for example, “Melbourne cir-

cus performance art schools”).

Table 4 shows the mean query lengths in char-

acters for each of the cohorts and animation condi-

tions. These means exclude one outlier, which was

a pasted query of length 61 (‘bachelor degree in cir-

cus arts that is available in Melbourne’) formulated

by a non-autistic user with the fast-rotating animation

condition. With the outlier included, the mean is 32.2.

The trend shows that in general, queries were longer

in the unanimated condition, with the difference in

ENASE 2022 - 17th International Conference on Evaluation of Novel Approaches to Software Engineering

456

query length being greater for non-autistic users.

Table 5 shows the proportion of queries that

were pasted for each cohort and condition. More

of those occurring in the animated condition were

pasted. Pasted queries tended to be longer (mean

query length was 37.9 characters, excluding the

outlier) than for unpasted (mean query length was

31.2 characters).

Table 5: Proportion of pasted queries.

Animation No Animation All

Autistic 0.14 0.00 0.08

Non-Autistic 0.31 0.23 0.27

Combined 0.25 0.17 0.21

4.1 Perceived Effort

After the task, participants answered a four-question

perceived effort questionnaire. Responses were on a

scale from 1 to 100 and varied greatly between partic-

ipants. These are summarised in Table 6.

Of note was that autistic participants under the

animation condition averaged 40.3 for the question

“How mentally demanding was the task”, whereas all

other cases had averages in the range 16.1 to 23.1.

For the question “How successful were you in accom-

plishing what you were asked to do?” autistic partic-

ipant responses averaged 76.3 for the animation case

and 71.2 in the unanimated case, whereas non-autistic

participants averaged 92.8 and 99.4 respectively.

For the question “How hard did you have to work

to accomplish your level of performance?”, non-

autistic participants in the unanimated case averaged

4.5 in their responses, whereas all other cases had av-

erages in the range 18.2 to 18.6. For the question

“How insecure discouraged irritated stressed and an-

noyed were you?”, both cohorts had similar scores in

the unanimated case (11.2 for autistic participants and

10 for non-autistic), but autistic participants scored

nearly double (37.9) that of non-autistic ones (20.9)

when animation was present.

4.2 Qualitative Data

The post-task survey question, “Could you please

type any further comments about the task”), elicited

a range of comments. The following codes were de-

ﬁned, using a grounded theory approach, by the ﬁrst

author and applied to their analysis:

• irritation/ distraction/ not distracting,

• avoidance,

• enjoyment,

• success/ feedback,

Table 6: Perceived effort.

Question Animation Autistic Mean

How mentally No No 16.1

demanding was No Yes 21.8

the task? Yes No 23.1

Yes Yes 40.3

How successful No No 99.4

were you in No Yes 71.2

accomplishing Yes No 92.8

what you were Yes Yes 76.3

asked to do?

How hard did you No No 4.5

have to work to No Yes 18.2

accomplish your Yes No 18.6

level of Yes Yes 18.3

performance?

How insecure, No No 10.0

discouraged, No Yes 11.2

irritated, stressed, Yes No 20.9

and annoyed Yes Yes 37.9

were you?

• easy/difﬁcult,

• confusion.

The third author also independently applied the codes,

leading to a Cohen’s kappa inter-coder agreement of

0.64. Combining the distraction code with the avoid-

ance code led to an agreement of 0.84. The main dif-

ference in coding was that two items were coded as

being about feedback by one author and difﬁculty and

confusion by the other.

Table 7 presents a summary of the qualitative anal-

ysis. We report below on the results of the analysis as

structured by the deﬁned codes.

Distraction. Eight participants (four autistic and

four non-autistic participants) commented on the

moving logo being distracting, for example, “The im-

age was too fast rotating; it signiﬁcantly distracted

me.” One non-autistic participant commented on the

layout colours being irritating in the unanimated con-

dition. Two non-autistic participants commented on

the unanimated image not distracting them.

Avoidance. Only one comment was associated with

this code: “I tried to complete this task quickly be-

cause of annoying rotating image.”. The comment

was provided by an autistic participant.

Enjoyment. Only one comment was associated

with this code: “enjoyed it”. The comment was pro-

vided by a non-autistic participant.

Success and Feedback. Two autistic users in the

unanimated case and one in the animated case com-

Web-based Search: How Do Animated User Interface Elements Affect Autistic and Non-autistic Users?

457

mented on it being “Hard to gauge success when no

search results came up.” Non-autistic users did not

comment on this. One in the unanimated case stated

“all was good”.

Difﬁculty. Two non-autistic and one autistic partici-

pant found the unanimated task easy; similarly for the

animated one.

Confusion. One autistic participant stated: “Easy

if I understood it correctly which I never know if I

have or not”.

Table 7: Summary of the qualitative analysis.

Code No. of comments with each code

Autistic Non-Autistic

irritation 1 (unanimated)

distraction 4 4

not distracting 2 (unanimated)

avoidance 1

enjoyment 1

feedback 3 1

difﬁculty 1 (easy) 2 (easy)

confusion 1

5 DISCUSSION

The case of search under the condition of irrelevant

animation has led to mixed results. Queries tended to

be formulated slightly faster and were more likely to

be shorter. Paradoxically, they were also more likely

to be pasted, which led to longer queries on aver-

age. The least successful query was typed under an

unanimated condition by an autistic participant. How-

ever, there was very little difference in query success

for this search task across conditions. This was, of

course, an artiﬁcial situation and the search task de-

scription was provided in a manner that allowed copy-

ing and pasting. Typed queries were shorter but not

slower overall. It would be interesting in a follow-up

to ﬁnd out the reasoning behind participants’ choice

of typing or pasting their query. It might be that

pasting was believed to be faster or possibly easier.

It is possible that those who typed their queries did

not think of an alternative method, or believed that it

wasn’t permitted for the study. Participants’ decisions

may have also depended on their typing skill.

In terms of how users felt during the task, autis-

tic users exposed to animation found it much more

mentally demanding than all other cases. Similarly,

they were far more irritated by animation than others.

Non-autistic users in the non-animated case found the

task the least arduous across all cases.

The question on success was answered differently

by the two cohorts, but this appears to be less related

to a difference in perceived success than a difference

in interpretation. The comments from autistic partici-

pants revealed a possible reason for this lower success

perception. Three autistic participants stated that the

lack of feedback after entering the query meant they

could not gauge their success.

6 LIMITATIONS

The purpose of the full study was to determine the

effect of on-screen animations on users, with partic-

ular focus on how this impacted autistic users, who

are known to have different responses to sensory in-

put than the general population. What is reported

here is on the search part of the overall study, and is

therefore not a comprehensive look at animation and

search. The participants did not receive feedback in

the form of search results and there was only a sin-

gle search task. The number of participants was small

with considerable variation in their responses, mean-

ing a purely statistical analysis would be unjustiﬁed

and extrapolating to all users would be risky.

It is possible that the differences between the

autistic and non-autistic cohorts would have been

greater if all participants had a similar language back-

ground. An artifact of recruitment was that the non-

autistic cohort were students, the majority of whom

did not have English as their ﬁrst language, whereas

this did not appear to be the case for the autistic co-

hort. Future studies are expected to account for this

variable. It should also be noted that the autistic par-

ticipants were without cognitive impairment (Autism

Spectrum Condition level 1) and that those at level 2

may be differently impacted.

7 CONCLUSIONS

In this paper we presented an analysis of the impact

of animated user interface elements on Web-based

search activities. The analysis examined how two

cohorts, autistic and non-autistic users, create search

queries for a speciﬁc search task in the presence or

absence of on-screen animations. Our aim was to an-

swer the following research question:

RQ: How does the effect of irrelevant animations

within user interfaces vary between autistic and non-

autistic users, while conducting Web search tasks?

Based on the results of our experiments, we can

conclude that there was little practical difference in

the time taken or the success of the query, but no-

ENASE 2022 - 17th International Conference on Evaluation of Novel Approaches to Software Engineering

458

table differences in strategy of query formulation,

with more people pasting their queries in the presence

of animation. In addition, autistic users found the task

more mentally demanding and irritating than non-

autistic users when animation was present. Thus, even

for short tasks such as formulating a search query, it

would be beneﬁcial to avoid having any irrelevant an-

imated elements in Web interfaces.

A possible future work direction is to replicate the

study with more queries and different types of anima-

tion, e.g., ﬂashing, because each type is likely to have

different effects on users, particularly autistic users.

Some types of animation, such as ﬂashing, are proba-

bly even more distracting than those used in our study,

and identiﬁcation of these potential issues might be

useful for making software more inclusive and acces-

sible.

ACKNOWLEDGEMENTS

The web application was adapted from a student pro-

gramming project.

We thank Gabrielle Hall for her input into the ex-

perimental design and insights into autistic research.

We also thank all participants for the time they in-

vested in our study.

REFERENCES

Alzahrani, M., Uitdenbogerd, A. L., and Spichkova, M.

(2021). Human-computer interaction: Inﬂuences on

autistic users. In Proceedings of the 25th International

Conference on Knowledge-Based and Intelligent In-

formation & Engineering Systems (KES’21).

Alzahrani, M., Uitdenbogerd, A. L., and Spichkova, M.

(2022). Impact of animated objects on autistic

and non-autistic users. In Proceedings of the 44th

International Conference on Software Engineering

(ICSE’22), to appear.

American Psychiatric Association (2013). Diagnostic and

statistical manual of mental disorders: DSM-5, vol-

ume 5. American psychiatric association Washington,

DC.

Battocchi, A., Ben-Sasson, A., Esposito, G., Gal, E., Pi-

anesi, F., Tomasini, D., Venuti, P., Weiss, P., and Zan-

canaro, M. (2010). Collaborative puzzle game: a

tabletop interface for fostering collaborative skills in

children with autism spectrum disorders. Journal of

Assistive Technologies.

Burke, M., Hornof, A., Nilsen, E., and Gorman, N. (2005).

High-cost banner blindness: Ads increase perceived

workload, hinder visual search, and are forgotten.

ACM T. Comput-Hum. INT., 12(4):423–445.

Cheung, M. Y., Hong, W., and Thong, J. Y. (2017). Effects

of animation on attentional resources of online con-

sumers. Journal of the Association for Information

Systems, 18(8):605–632.

C¸ orlu, D., Tas¸el, S¸., Turan, S. G., Gatos, A., and Yantac¸,

A. E. (2017). Involving autistics in user experience

studies: A critical review. In Proceedings of the 2017

Conference on Designing Interactive Systems, pages

43–55, Edinburgh, United Kingdom.

Eraslan, S., Yaneva, V., Yesilada, Y., and Harper, S. (2019).

Web users with autism: eye tracking evidence for dif-

ferences. Behav. Inform. Technol., 38(7):678–700.

Fisher, R. A. and Yates, F. (1938). Statistical tables: For

biological, agricultural and medical research. Oliver

and Boyd.

Gentry, T., Wallace, J., Kvarfordt, C., and Lynch, K. B.

(2010). Personal digital assistants as cognitive aids

for high school students with autism: Results of a

community-based trial. Journal of Vocational Reha-

bilitation, 32(2):101–107.

Happ

e, F. and Frith, U. (2006). The weak coherence ac-

count: detail-focused cognitive style in autism spec-

trum disorders. Journal of autism and developmental

disorders, 36(1):5–25.

Hong, W., Thong, J. Y., and Tam, K. Y. (2007). How do

Web users respond to non-banner-ads animation? The

effects of task type and user experience. .J AM. SOC.

INF. SCI. TEC., 58(10):1467–1482.

Lohtia, R., Donthu, N., and Hershberger, E. K. (2003). The

impact of content and design elements on banner ad-

vertising click-through rates. Journal of advertising

Research, 43(4):410–418.

Majrashi, K. (2016). Cross-platform user experience. PhD

thesis, RMIT University.

Millen, L., Edlin-White, R., and Cobb, S. (2010). The de-

velopment of educational collaborative virtual envi-

ronments for children with autism. In Proceedings of

the 5th Cambridge Workshop on Universal Access and

Assistive Technology, Cambridge, volume 1, page 7.

Pagendarm, M. and Schaumburg, H. (2001). Why are users

banner-blind? the impact of navigation style on the

perception of web banners. Journal of Digital Infor-

mation, 2(1):14.

Purcell, K. (2011). Search and email still top the list of most

popular online activities. Pew Internet & American

Life Project Washington, DC.

Putnam, C. and Chong, L. (2008). Software and technolo-

gies designed for people with autism: what do users

want? In Proceedings of the 10th international ACM

SIGACCESS conference on Computers and accessi-

bility, pages 3–10.

Rau, P.-L. P., Chen, J., and Chen, D. (2006). A study of pre-

sentations of mobile web banners for location-based

information and entertainment information websites.

Behaviour & Information Technology, 25(03):253–

261.

Schaaf, R. C. and Miller, L. J. (2005). Occupational ther-

apy using a sensory integrative approach for chil-

dren with developmental disabilities. Mental retarda-

Web-based Search: How Do Animated User Interface Elements Affect Autistic and Non-autistic Users?

459

tion and developmental disabilities research reviews,

11(2):143–148.

Seeman, L. and Cooper, M. (2015). Cognitive accessibility

user research. W3C First Public Working Draft, 15.

Sitdhisanguan, K., Chotikakamthorn, N., Dechaboon, A.,

and Out, P. (2012). Using tangible user interfaces in

computer-based training systems for low-functioning

autistic children. Personal and Ubiquitous Comput-

ing, 16(2):143–155.

Valencia, K., Rusu, C., Qui

nones, D., and Jamet, E. (2019).

The Impact of Technology on People with Autism

Spectrum Disorder: A Systematic Literature Review.

Sensors, 19(20):4485.

Vos, T., Allen, C., Arora, M., Barber, R. M., Bhutta, Z. A.,

Brown, A., Carter, A., Casey, D. C., Charlson, F. J.,

Chen, A. Z., et al. (2016). Global, regional, and na-

tional incidence, prevalence, and years lived with dis-

ability for 310 diseases and injuries, 1990–2015: a

systematic analysis for the global burden of disease

study 2015. The lancet, 388(10053):1545–1602.

Wang, Q., Yang, S., Liu, M., Cao, Z., and Ma, Q. (2014).

An eye-tracking study of website complexity from

cognitive load perspective. Decision support systems,

62:1–10.

Yesilada, Y. and Harper, S., editors (2019). Web Accessi-

bility - A Foundation for Research, Second Edition.

Human-Computer Interaction Series. Springer.

Yoo, C. Y., Kim, K., and Stout, P. A. (2004). Assessing

the effects of animation in online banner advertising:

Hierarchy of effects model. Journal of interactive ad-

vertising, 4(2):49–60.

Zhang, Y. and Moffat, A. (2006). Some observations on

user search behaviour. Australian Journal of Intelli-

gent Information Processing Systems, 9(2):1–8.

Zorn, S., Olaru, D., Veheim, T., Zhao, S., and Murphy, J.

(2012). Impact of animation and language on banner

click-through rates. Journal of Electronic Commerce

Research, 13(2):173–183.

ENASE 2022 - 17th International Conference on Evaluation of Novel Approaches to Software Engineering

460