The Usability of Hidden Functional Elements in Mobile User

Interfaces

Mubashar Munir and Pietro Murano

Department of Computer Science, Oslo Metropolitan University, Postboks 4, St. Olavs plass 0130, Oslo, Norway

Keywords: Usability, Hidden Elements, Mobile User Interfaces, Evaluation.

Abstract: The trend of maximizing mobile screen real estate by hiding user interface features has been in use for some

time. However, there is a lack of empirical knowledge concerning the real usability issues of using this

strategy. In this paper, we present novel and statistically significant evidence to suggest that hiding user

interface elements decreases usability in terms of performance and user experience. We conducted a within-

users experiment comparing identical user interfaces, where the only differences between them were that one

version hid the user interface elements and the other version had all the user interface elements visible to the

user. We recorded task times, errors and user satisfaction for a series of tasks. We also discuss our results in

light of existing user interface design guidelines and show that our results are in harmony with existing

guidelines.

1 INTRODUCTION

In this paper we present the results of an investigation

into the common trend of hiding functional elements

at the user interface for increasing screen real estate

on mobile devices such as smartphones and tablets.

Such hidden elements are typically accessed from

within an app by swiping one of the sides of the

mobile device and then the extra features available in

an app appear on the screen. When not in use, or if

one swipes in the opposite direction of access, these

disappear again from the user interface.

While this solution typically maximises the screen

real estate available, it reduces the overall usability of

an app, because users need to remember where the

features are and are required to make more interaction

touches to interact with the app. Frequent use of such

apps will likely ensure users can remember how to

access such features in the future. However, this may

not be the case for infrequent use of such an app.

Also, forcing users to swipe from a side in order to

access other app features could increase task times

and errors while also reducing user satisfaction.

Therefore, the authors of this paper investigated

the issue of hiding features within an app with a view

to maximizing screen real estate. This was done

specifically in relation to effects on task time, errors

and user satisfaction. To our knowledge this direct

and specific investigation has not been done before

and closes an important gap in knowledge.

In the following sections we discuss some

background literature related to this investigation.

Then our experiment and results are presented.

Finally, some discussions and conclusions conclude

the paper.

2 BACKGROUND

The issue of screen real estate available on a mobile

device has been studied for some time. Clearly,

smaller screens incur interaction challenges and

therefore researchers have tried many strategies to

overcome the issue of small screens or less screen real

estate.

Gomes, Priyadarshana, Visser, Carrascal and

Vertegaal (2018) explored the area of using flexible

displays that are held in a roll shape, but which can

then be unrolled to reveal a tablet-sized screen, thus

giving more screen-real estate than with most

conventional smartphones. The authors concluded

that tasks involving navigation were intuitive with

their prototype device. Their set of participants

seemed to indicate that the shape of their device could

improve grasping when compared with a telephone

shape.

256

Munir, M. and Murano, P.

The Usability of Hidden Functional Elements in Mobile User Interfaces.

DOI: 10.5220/0011827700003467

In Proceedings of the 25th International Conference on Enterprise Information Systems (ICEIS 2023) - Volume 2, pages 256-263

ISBN: 978-989-758-648-4; ISSN: 2184-4992

 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

Concerning small screens and touch interaction,

Butler, Izadi and Hodges (2008) tried to look at the

issue of the fingers potentially occluding some of the

already small screen area on most mobile devices

during an interaction. Their approach was to trial

‘infra-red proximity sensors embedded along each

side’ (Butler et al, 2008) of the mobile device. This

enabled detection and positioning of the fingers in

nearby areas of the screen, effectively increasing the

interaction zone size. Although rigorous evaluation

was not reported by the authors, the concept is

interesting.

Further, Song, Sörös, Pece, Fanello, Izadi, Keskin

and Hilliges (2014) developed an algorithm which

would recognize in-air gestures using only the

standard camera found on a mobile device. The idea

was to maximise the screen real estate available by

including the possibility of in-air gestures in the

proximity of the mobile device. This would also

somewhat alleviate issues of finger occlusion during

interaction. Their comparative evaluation consisting

of touch interaction only versus touch and in-air

gestures, suggested that interacting with a mobile

device where touch and in-air gestures were available

simultaneously is statistically significantly faster.

Informal evaluation with a few individuals also

suggested that users were mostly positive towards

using in-air gestures for interaction.

Byrd and Caldwell (2009) compared 2.8”, 3.5”

and 7” screen sizes. While not all results produced

significant outcomes, the overall trend showed that

larger screen sizes produced faster task times and

shorter screen access times. Differences in errors for

each screen size were not statistically significant.

This gave users a larger input area, particularly if the

device rested on a surface.

Also, Wang, Hsieh and Paepcke (2009)

investigated the option of deliberately hiding files (or

content) out of sight on a small screen device.

Content, such as photographs, etc. could be placed in

virtual ‘piles’ away from the main screen real estate,

but available by accessing the sides of the screen,

through small visual cues located at the sides of the

screen. These represented distinct ‘piles’ of content,

e.g. photographs or categories of photographs, etc.

Although the concept suggested is interesting, it

forces users to rely on their memory. It also raises

doubts concerning the length of time it would take

users to forget (or partially forget) what each ‘pile’

contains. There is also lack of evidence concerning

the optimum amount of ‘piles’ a user should use,

before increased quantities of ‘piles’ would become

easier to forget.

This brief consideration of previous works shows

that a lot of effort has been expended over the years

in trying to deal with the basic issue of mobile devices

having very limited screen real estate.

However, the main solution that has been adopted

in recent years is to hide as many of the user interface

elements as possible so as to maximise the use of the

physical dimensions available of mobile devices.

However, as will be seen from the evidence presented

below, this is not the best option in terms of usability.

Therefore, in the next section we present detailed

information on the experiment we conducted in

relation to hidden user interface elements.

3 EXPERIMENT

3.1 Experiment Introduction

In order to evaluate the usability of hiding functional

elements on a mobile user interface it was decided to

use a mobile camera-type application as the main

context. The mobile camera was chosen for several

reasons. (1) As far as we know all new or not so old

smartphones include one or more cameras. (2) The

camera application on a smartphone is one which

often contains several functionalities that can be

tailored or adopted by users and is therefore suitable

for an experiment and a series of realistic tasks which

would be ecologically valid. (3) Camera usage on

smartphones is very common amongst users. Already

in 2017, one survey indicated that 85% of all digital

photographs were taken with a smartphone (Richter,

2017).

We chose to use an empirical experiment with

hypotheses (Lazar, Feng and Hochheiser, 2017),

rooted in the hypothetico-deductive approach.

However, our experiment also contained a more

qualitative side to it, by using a questionnaire, with a

Likert-type scale (Likert, 1932) to elicit participant

opinions. The main reasons for this, were that we

wanted to conduct the study in controlled conditions

in order to collect precise numerical data, whilst at the

same time collecting some qualitative data. We felt

this approach was more useful than having a

completely quantitative or completely qualitative

approach. Furthermore, this is an approach that has

been used in previous research to good effect (e.g.

Shrestha and Murano (2022), Keya and Murano

(2022) and Shrestha and Murano (2016)).

The Usability of Hidden Functional Elements in Mobile User Interfaces

257

3.2 Experiment Hypotheses

In our design of the experiment, we devised a series

of two-tailed hypotheses. These are as follows:

H1 - There will be a statistically significant difference

between the two developed prototypes in terms of

task times.

– There will be no statistically significant

difference between the two developed prototypes in

terms of task times.

H2 - There will be a statistically significant difference

between the two developed prototypes in terms of

user errors.

- There will be no statistically significant

difference between the two developed prototypes in

terms of user errors.

H3 - There will be a statistically significant difference

between the two developed prototypes in terms of

user satisfaction.

- There will be no statistically significant

difference between the two developed prototypes in

terms of user satisfaction.

3.3 Users

In this experiment 20 participants were recruited

within varied age ranges (21-30 years – 9 participants,

31-40 years – 6 participants, 41-50 years – 3

participants, 51-60 years – 2 participants). Our

sample of participants consisted of 10 males and 10

females.

All the participants had in common that they were

familiar in advance in using a camera application.

Overall, all participants were competent in the use of

mobile devices/computers.

These common aspects of knowledge amongst the

sample ensured that the overall sample was relatively

homogeneous. This was important especially as some

of the data collected could have been biased by overly

novice or overly experienced participants, e.g. the

timings and errors recorded.

All ethical considerations concerning human

participants were implemented in line with

Norwegian guidelines for such research.

3.4 Design

For this experiment, a within-users experimental

design was chosen. The main reason for this choice

was that it allowed each participant to experience

both user interface types and thus evaluate their

experiences based on having interacted with two

different user interfaces.

3.5 Variables

The independent variables were the two user

interfaces being evaluated and the specific tasks used

in connection with the user interfaces.

The dependent variables were in connection with

performance and user satisfaction.

The dependent measures were the time to

complete a task and the number of errors. Errors were

simply defined as a participant not completing a task.

A post-experiment questionnaire with six specific

user experience-type questions was used to reveal

some details on user satisfaction.

3.6 Tasks

The tasks were designed to be as realistic as possible

within a camera application. The five tasks we

designed were: To capture a picture, To adjust the

"Zoom Level", To turn off the "Flash", To open the

"Settings" and To adjust the "Auto White Balance"

AWB. Therefore, each prototype had five working

options – one for each task. During the experiment the

tasks were executed by participants in the same order

as listed here.

The appearance of the two prototypes used is

shown in Figure 1.

3.7 Procedure

The experiment was conducted in three main stages.

The first stage welcomed the participants and at this

point participants signed a consent form (Note: the

experiment met ethical standards for Norway) in

relation to having read an information sheet

concerning the research.

The second stage consisted of carrying out the

tasks, as described in the previous section. During this

stage the participants were observed for their

interactions and the results of each task were recorded

on a spreadsheet.

The third and final stage involved the participants

completing a post-experiment questionnaire, which

elicited opinions concerning the experiences they had

had in using the two different prototypes.

The next section will present the results of the

statistical analysis on the collected data.

ICEIS 2023 - 25th International Conference on Enterprise Information Systems

258

Figure 1: The Final Design of the Prototypes.

4 RESULTS

In this section we present the results of a statistical

analysis on the data collected and described earlier in

this paper. The main aim was to determine the

presence of any statistical significance.

The data was initially analysed using the Shapiro-

Wilk test, which indicated that the data was

parametric in nature and therefore suitable to be

analysed further with a parametric significance test.

For the data we chose a t-test. We have chosen to

show in this section the precise figures for the means,

standard deviations, t-tests and p values for readers

interested in the details. However, we also present the

means and standard deviations in graphical format to

allow readers who wish only a quick overview to

achieve this.

Task Time: Overall task time (minutes and

seconds) was recorded as the total time taken for all

tasks on each prototype by a participant.

Descriptive statistics show that the Mean (M) task

time in using Prototype 1 with the hidden elements is

M = 1.1470, Standard Deviation (SD) = .2215 and the

Mean task time in using Prototype 2 with the visible

elements is M = .3095, SD = .0697. The t-test result

is t = 18.142, p < .001. This shows a highly significant

difference in the overall task times, where the

prototype with visible elements was significantly

faster to use to complete the designated tasks. Figure

2 displays the means and standard deviations for the

task times.

Figure 2: Means and Standard Deviations for Task Times.

Number of Errors: As discussed above, errors

were defined as a failure to complete a task. One error

in performing a task was counted as a failure. A total

of 15 errors were observed for prototype 1 with

hidden elements, and there were no errors observed

for prototype 2 with visible elements.

Descriptive statistics show that the Mean (M)

number of errors incurred in using Prototype 1 with

the hidden elements is M = .75, SD = .550 and the

Mean number of errors incurred in using Prototype 2

with the visible elements is M = 0, SD = 0. The t-test

result is t = 6.097, p < .001. This shows a highly

significant difference in the overall error rate, where

the prototype with visible elements was significantly

better at helping users avoid making errors. Figure 3

displays the means and standard deviations for the

errors made during the tasks.

1,147

0,3095

0,2215

0,0697

PROTOTYPE HIDDEN

ELEMENTS

PROTOTYPE VISIBLE

ELEMENTS

TASK TIMES

Mean Standard Deviation

The Usability of Hidden Functional Elements in Mobile User Interfaces

259

Figure 3: Means and Standard Deviations for the Errors

Made During the Tasks.

User Satisfaction: In order to find out

participants’ opinions regarding user satisfaction, a

seven-question questionnaire was designed. Each

question was asked for each user interface version

and the scores for each pair were statistically analysed

for significant differences. Each question in the

questionnaire used a Likert-type scale ranging from 1

to 5 (Strongly Disagree = 1, Disagree = 2, Neutral =

3, Agree = 4, Strongly Agree = 5). Higher scores

indicated higher user satisfaction.

Question 1 concerned the user interface being

easy to interact with. Descriptive statistics show that

the Mean (M) response for Prototype 1 with the

hidden elements is M = 2.60, SD = 1.353 and the

Mean response for Prototype 2 with the visible

elements is M = 4.40, SD = 1.273. The t-test result is

t = -3.488, p = .002. This shows a statistically

significant difference in overall opinions, where the

prototype with visible elements was rated

significantly easier to interact with. Figure 4 displays

the means and standard deviations for the user

interface being easy to interact with.

Figure 4: Means and Standard Deviations for the User

Interface Being Easy to Interact With.

Question 2 concerned how easy it was to

remember the commands. Descriptive statistics show

that the Mean (M) response for Prototype 1 with the

hidden elements is M = 2.45, SD = 1.356 and the

Mean response for Prototype 2 with the visible

elements is M = 3.85, SD = 1.226. The t-test result is

t = -2.692, p = .01. This shows a statistically

significant difference in overall opinions, where the

prototype with visible elements was rated

significantly easier for one to remember the

commands. Figure 5 displays the means and standard

deviations for the ease of remembering the

commands.

Figure 5: Means and Standard Deviations for the Ease of

Remembering the Commands.

Question 3 concerned the tasks being easy to

complete on the interface. Descriptive statistics show

that the Mean (M) response for Prototype 1 with the

hidden elements is M = 3.25, SD = 1.209 and the

Mean response for Prototype 2 with the visible

elements is M = 4.50, SD = 1.147. The t-test result is

Figure 6: Means and Standard Deviations for the Ease of

Completing the Tasks at the User Interface.

0,75

0,55

PROTOTYPE HIDDEN

ELEMENTS

PROTOTYPE VISIBLE

ELEMENTS

ERRORS

Mean Standard Deviation

2,6

4,4

1,353

1,273

PROTOTYPE HIDDEN

ELEMENTS

PROTOTYPE VISIBLE

ELEMENTS

USER INTERFACE EASY TO

INTERACT WITH

Mean Standard Deviation

2,45

3,85

1,356

1,226

PROTOTYPE HIDDEN

ELEMENTS

PROTOTYPE VISIBLE

ELEMENTS

EASE OF REMEMBERING

COMMANDS

Mean Standard Deviation

3,25

4,5

1,209

1,147

PROTOTYPE HIDDEN

ELEMENTS

PROTOTYPE VISIBLE

ELEMENTS

EASE OF COMPLETING THE

TASKS WITH THE USER

INTERFACE

Mean Standard Deviation

ICEIS 2023 - 25th International Conference on Enterprise Information Systems

260

t = -3.101, p = .006. This shows a statistically

significant difference in overall opinions, where the

prototype with visible elements was rated

significantly easier for the tasks used in the

evaluation. Figure 6 displays the means and standard

deviations for the ease of completing the tasks at the

user interface.

Question 4 concerned the clarity of one finding a

desired option/command. Descriptive statistics show

that the Mean (M) response for Prototype 1 with the

hidden elements is M = 2.30, SD = 1.302 and the

Mean response for Prototype 2 with the visible

elements is M = 4.10, SD = 1.447. The t-test result is

t = -3.214, p = .005. This shows a statistically

significant difference in overall opinions, where the

prototype with visible elements was rated

significantly clearer in terms of finding a particular

option or command. Figure 7 displays the means and

standard deviations for the clarity in finding a desired

option/command.

Figure 7: Means and Standard Deviations for the Clarity in

Finding a Desired Option/Command.

Question 5 concerned feelings of comfort with

hidden elements or visible elements. Descriptive

statistics show that the Mean (M) response for

Prototype 1 with the hidden elements is M = 2.75, SD

= 1.164 and the Mean response for Prototype 2 with

the visible elements is M = 3.85, SD = 1.089. The t-

test result is t = -2.463, p = .02. This shows a

statistically significant difference in overall opinions,

where the prototype with visible elements was rated

as significantly more comfortable in featuring visible

user interface elements. Figure 8 displays the means

and standard deviations for the feelings of comfort

with hidden elements or visible elements.

Figure 8: Means and Standard Deviations for the Feelings

of Comfort With Hidden Elements or Visible Elements.

Question 6 concerned feelings of comfort with

the app interface using maximum screen space by

hiding elements or the app interface occupying more

screen space for displaying commands/options.

Descriptive statistics show that the Mean (M)

response for Prototype 1 with the hidden elements is

M = 2.75, SD = 1.070 and the Mean response for

Prototype 2 with the visible elements is M = 3.55, SD

= .999. The t-test result is t = -2.270, p = .03. This

shows a statistically significant difference in overall

opinions, where the prototype with visible elements

was rated as significantly more comfortable despite

having less screen real estate available due to

displaying options/commands (rather than hiding

them). Figure 9 displays the means and standard

deviations for the feelings of comfort with the app

interface screen space.

Figure 9: Means and Standard Deviations for the Feelings

of Comfort With the App Interface Screen Space.

2,3

4,1

1,302

1,447

PROTOTYPE HIDDEN

ELEMENTS

PROTOTYPE VISIBLE

ELEMENTS

CLARITY IN FINDING AN

OPTION/COMMAND

Mean Standard Deviation

2,75

3,85

1,164

1,089

PROTOTYPE HIDDEN

ELEMENTS

PROTOTYPE VISIBLE

ELEMENTS

FEELINGS OF COMFORT WITH

HIDDEN/VISIBLE ELEMENTS

Mean Standard Deviation

2,75

3,55

1,07

0,999

PROTOTYPE HIDDEN

ELEMENTS

PROTOTYPE VISIBLE

ELEMENTS

FEELINGS OF COMFORT WITH

THE APP INTERFACE SCREEN

SPACE

Mean Standard Deviation

The Usability of Hidden Functional Elements in Mobile User Interfaces

261

Question 7 asked participants to make one choice

regarding which of the two user interface types they

would prefer to use. In line with the responses of the

previous six questions participants overwhelmingly

(80%) responded that they would choose the user

interface where the user interface elements were

displayed (not hidden). The other 20% of the sample

would choose the user interface where the elements

were hidden from view.

This section has presented detailed results in

terms of performance in carrying out representative

tasks and user opinions. In the next section we discuss

how real-world use of our findings would improve the

usability of mobile interaction.

5 DISCUSSION

The results detailed in the previous section are quite

clear and categorical. We have novel and significant

results. All the performance and user perception

aspects we investigated show clearly that hidden user

interface elements perform much more poorly and are

strongly disliked when compared with an equivalent

counterpart that does not hide any of the elements. To

our knowledge no other studies have directly and

specifically investigated by direct empirical

comparison the hiding of user interface elements for

increasing screen real estate. Further, the results

should be a clear signal to all user interface designers

that hiding user interface elements is not the best way

forward.

Therefore, our three positive hypotheses (see

Experiment Hypotheses section above) which related

to task times, user errors and user satisfaction are all

accepted. In each of the three cases we observed,

categorically statistically significant figures to

suggest that in all three areas under investigation the

visible elements perform significantly better and are

significantly preferred over a version of a user

interface designed to have invisible elements.

The results of this investigation are also very

much in line with current knowledge of user interface

design. For example, Nielsen’s Heuristics (Nielsen,

2020), which can be used in an evaluative process

and/or a design process suggest in Heuristic 6 to use

‘Recognition rather than recall’ (Nielsen, 2020). In

further explaining this, we are encouraged to

‘Minimize the user's memory load by making

elements, actions, and options visible. The user

should not have to remember information from one

part of the interface to another. Information required

to use the design (e.g. field labels or menu items)

should be visible or easily retrievable when needed.

(Nielsen, 2020) (see also Budiu, (2014))’ This

suggests that hiding necessary elements of an app in

order to have more screen real estate is not the best

option as it forces users to remember where

everything is.

Furthermore, the International Organization for

Standardization (ISO) provides support for the

findings in this investigation and Nielsen’s

Heuristics. The ISO 9241-110:2021 (International

Organization for Standardization, 2021) concerns

Interaction Principles. Sub-section 5.4.1 under the

Learnability section outlines various principles

related to Discovery. One of these is that ‘the

interactive system supports discovery of its

capabilities and how to use them, allows exploration

of the interactive system, minimizes the need for

learning and provides support when learning is

needed. (International Organization for

Standardization, 2021)’. We would suggest that

hiding user interface features, etc. does not support

discovery or learning.

6 CONCLUSIONS

The trend over recent years to maximise at all costs

the screen real estate for mobile devices is probably

unnecessary, as suggested by our investigation and

well-established guidelines and principles. Even

though human users are generally good at adapting to

different modes of interaction, e.g. in remembering

where hidden options are within an app, we would

argue that this does not justify bad usability practices

and bad user interface design.

It is acknowledged that there needs to be a trade-

off between screen real estate and making all

elements (e.g. app features, settings and/or navigation

etc.) visible. However, the trend to hide everything is

not the best option given our investigation results and

the well-established user interface design principles.

However, more and better design decisions should

lead designers to not hiding everything, but to making

at least the most important aspects visible. This

clearly suggests that some elements would continue

to be hidden. If that would be the case, better user

interface cues that do not use much screen space

could be adopted to let users know there are other

options available.

Future work in this area would benefit by

investigating the option of using user interface cues

to inform a user that other elements are available.

However, our suggestion is that if the number of

elements are not too numerous, these should always

be visible by default, with perhaps an easy option to

ICEIS 2023 - 25th International Conference on Enterprise Information Systems

262

make them invisible for experienced users. Further,

this investigation did not specifically look at the

effects on human memory (short or long term

memory) in relation to hiding/making visible user

interface elements. Future work could also focus on

this aspect and provide useful insights into these

approaches.

REFERENCES

Budiu, R. (2014) Memory Recognition and Recall in User

Interfaces,

https://www.nngroup.com/articles/recognition-and-

recall/

Butler, A., Izadi, S. and Hodges, S. (2008) SideSight:

Multi-“touch” Interaction Around Small Devices, UIST

'08: Proceedings of the 21st annual ACM symposium

on User interface software and technology, October

2008, P. 201–204.

Byrd, K.S and Caldwell, B.S. (2009) From UMPCs to Cell

Phones: How Does Diminishing Screen Real Estate

Affect Screen Access and Working Memory?,

Proceedings of the Human Factors and Ergonomics

Society 53rd Annual Meeting.

Gomes, A., Priyadarshana, L.L., Visser, A., Carrascal, J.P.

and Vertegaal, R. (2018) MagicScroll: A Rollable

Display Device With Flexible Screen Real Estate and

Gestural Input, MobileHCI'18, September 3-6,

Barcelona, Spain, ACM.

International Organization for Standardization (2021)

Ergonomics of Human-System Interaction - Part 110:

Interaction Principles (ISO 9241-110:2021).

Keya, R. T. and Murano, P. (2022) Is Carousel Interaction

Really Usable?, First Monday, Volume 27, Issue 1,

January 2022.

Lazar, J., Feng, J. H. and Hochheiser, H. (2017). Research

Methods in Human-Computer Interaction, Morgan

Kaufmann.

Likert, R. (1932) A technique for the measurement of

attitudes. Archives of psychology, number 140. New

York: Science Press.

Nielsen, J. (2020) 10 Usability Heuristics for User Interface

Design, https://www.nngroup.com/articles/ten-usabi

lity-heuristics/

Richter, F. (2017) Smartphones Cause Photography Boom,

Statista, Aug 31, https://www.statista.com/chart/10913/

number-of-photos-taken-worldwide/

Shrestha, S. L and Murano. (2022) The Design and

Evaluation of an Online Dictionary User Interface -

PDF, International Journal of Computing and Digital

Systems, Volume 1, March 2022.

Shrestha, S. and Murano, P. (2016). A Universally

Designed and Usable Data Visualization for A Mobile

Application in the Context of Rheumatoid Arthritis.

International Journal of Advanced Computer Science

and Applications, Vol. 7, Issue 12, December.

Song, J., Sörös, G., Pece, F., Fanello, S.R., Izadi, S.,

Keskin, C. and Hilliges, O. (2014) In-air Gestures

Around Unmodified Mobile Devices, UIST '14:

Proceedings of the 27th annual ACM symposium on

User interface software and technology, October 2014,

P. 319–329.

Wang, Q.Y., Hsieh, T. and Paepcke, A. (2009) Piles Across

Space: Breaking the Real-Estate Barrier on Small-

Display Devices, International Journal of Human-

Computer Studies 67 (2009) 349–365.

The Usability of Hidden Functional Elements in Mobile User Interfaces

263