User Time Spent between Persuasiveness and Usability of Social

Networking Mobile Applications: Patterns of Influence

Mohammed Bedjaoui, Nadia Elouali and Sidi Mohamed Benslimane

LabRI-SBA Lab., Ecole Superieure en Informatique, Sidi Bel Abbes, Algeria

Keywords: Persuasive Technology, Usability, Users’ Time Spent.

Abstract: Using social media is one of the most common activities for mobile users. Moreover, it is a time-consuming

activity that can lead to addiction. Some gaps in HCI (Human Computer Interaction) ergonomics theory gave

rise to this addiction. These gaps lie in an overexploitation of the usability and/or persuasion criteria that

designers and/or developers use according to their needs when applying influence strategies to affect users’

engagement. Although these strategies are widely applied in online social networks, they are not well

identified and their application levels are still lacking. This paper seeks to establish and validate these

influence strategies. We proposed five (05) patterns of influence in online social networks that have a

significant impact on Users’ Time Spent (UTS) grouping the different usability criteria and persuasion

strategies. Then, we conducted a classification study of those criteria / strategies, using Hybrid Card Sort

method carried out by fifteen (15) eligible experts. Experts were asked to group those criteria / strategies into

a set of patterns based on our predetermined (with the option to create their own patterns). The results analysis

validates our five (05) proposed patterns paving the way to outline their application borderline thereafter.

1 INTRODUCTION

The use of social media such as Facebook, Instagram

and YouTube has become the cornerstone of modern

communication and connection

. The average social

media user spends 135 minutes a day online.

Knowing that 60 percent of social media consumption

comes from mobile devices

A new purpose has emerged, with the great

evolution of social media: to capture the attention of

users and to maximize their engagement. In other

words, maximize the time that users spend with social

media, which can lead to addiction (Kuss and

Griffiths, 2011).

In our paper (Bedjaoui et al., 2018), we presented

our study about the usability and persuasion of the

most used social networking interfaces Facebook and

YouTube according to the UTS (Users' Time Spent).

We conducted heuristic inspection using (Bastien and

Scapin, 1993) usability heuristics and (Némery and

Brangier, 2014) persuasion criteria grid to evaluate

both interfaces. We showed that there are some gaps

in the HCI (Human-Computer Interaction)

https://www.psychologytoday.com/us/blog/in-

excess/201805/addicted-social-media

ergonomics theory that gave rise to addictive

interfaces. Some usability criteria (Guidance, Explicit

control, Compatibility) and persuasion criteria

(Personalization, Priming, Commitment,

Ascendency) are overexploited for the purpose of

persuading users to spend as much time as possible.

(The “overexploitation” of persuasive techniques and

ergonomic criteria, turns the techniques that should

serve the user into techniques that manipulate

him/her).

These overexploitation lie in the fact that

designers and/or developers use their judgments to

implement the different usability and persuasion

criteria. In addition, these criteria have no application

borderline. Thus, designers or developers can apply

them according to their needs or interests without

necessarily considering the users objectives (how

they want to spend their time), their relationships

(interpersonal and social choices), their mental and

emotional limits (vulnerabilities, tiredness and ways

in which their minds form habits), etc.

In the present paper, we seek to frame the various

techniques and strategies of influence used by social

media that aim to persuade users to stay online for as

https://strikesocial.com/blog/50-social-media-stats-every-

marketer-should-know-in-2018/

Bedjaoui, M., Elouali, N. and Benslimane, S.

User Time Spent between Persuasiveness and Usability of Social Networking Mobile Applications: Patterns of Inﬂuence.

DOI: 10.5220/0008976400350044

In Proceedings of the 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2020) - Volume 2: HUCAPP, pages

35-44

ISBN: 978-989-758-402-2; ISSN: 2184-4321

long as possible and that led to overexploitation of

usability / persuasion criteria, into a set of patterns.

Our purpose is to establish the influence patterns

in order to define their application borderline

thereafter.

At first, we proposed five (05) patterns that

regroup, the different usability criteria and persuasion

strategies (Suggestion Pattern, Reminder Pattern,

Reward Pattern, Interaction Pattern and Social

Influence Pattern). Then and in order to validate our

proposed patterns, we conducted a classification

study of those criteria / strategies, using Hybrid Card

Sort method carried out by fifteen (15) eligible

experts.

The remainder of the paper is organized as

follows. In section 2 we give a background of the

usability and persuasive technology. In section 3, we

discuss related work. We give details on the study

methodology in section 4, and we present the results

and their analyses in section 5. In section 6 we discuss

the application and relationship between these

patterns as well as their influence on UTS. Finally,

section 7 presents some concluding remarks and

outlines topics for future research.

2 THEORETICAL

BACKGROUND

2.1 Usability

The concept of usability aims to develop simple and

effective interactions that take into account users’

reasoning, their workload, their constraints

(environment, position, and task) and strain (fatigue).

Several definitions have been proposed for

usability, the well-known and commonly used was

proposed by (ISO 9241-11, 1998): “The extent to

which a system, product or service can be used by

specified users to achieve specified goals with

effectiveness, efficiency and satisfaction in a

specified context of use”.

Other definitions of usability have been proposed.

They focused on more specific usability criteria, for

example, (Nielsen, 1994) considered “efficiency”,

“ease of learning”, “memorization”, “errors”,

“safety” and “satisfaction”, where (Shneiderman et

al., 2016) emphasized “efficiency” and

“effectiveness”.

Several grids of usability criteria have been

proposed for the evaluation and design of products

and systems. The best known are those of (Bastien

and Scapin, 1993) including several elementary

criteria grouped into eight general principles:

Guidance: All the means implemented to advise,

direct, inform and lead the user during his interactions

(messages, alarms, labels, etc.)

Workload: A set of interface elements that help

reduce user’s perception and memorization load and

increase the dialogue effectiveness.

Explicit Control: Concerns both the system

processing of the explicit actions of users and the

control that users have over the handling of their

actions by the system.

Adaptability: Refers to the ability of a system to

react according to the context, and according to the

needs and preferences of users.

Error management: All the means that aim to avoid

or reduce errors, and to correct them when they occur.

Consistency: Refers to the way in which interface

design choices (codes, denominations, formats,

procedures, etc.) are kept for the same contexts, and

are different for different contexts.

Significance of Codes: Refers to the adequacy

between the object or information displayed or input,

and its referent.

Compatibility: Refers to the agreement that can exist

between the characteristics of the users and the tasks,

on one hand, and the organization of the outputs, the

inputs and the dialogue of a given application, on the

other hand.

2.2 Persuasive Technology

Persuasive design, persuasive interfaces, persuasive

technology and Captology are all denominations,

which, according to the literature refer to the same

phenomenon. All of these terminologies refer to

modifying (or attempting to change) the attitude or

behaviour of users through interfaces and guiding

them to take actions that they would not have done

spontaneously. (Fogg, 2002) has been among the first

contributors in the field of persuasion technology for

which he coined the term “Captology,” an

abbreviation for “Computers As Persuasive

Technology”, which he defines as “design, research,

and analysis of interactive computing products

created for the purpose of changing people’s attitudes

or behaviours”. This concept describes an area in

which technology and persuasion overlap. (As shown

in Figure 1.)

HUCAPP 2020 - 4th International Conference on Human Computer Interaction Theory and Applications

Figure 1: Captology.

He introduced the six basic levers of persuasive

technology:

 Computer as Persuasive Tools: Presents the

persuasive technologies that aim to bring users to

the expected behaviour or attitude by increasing

their ability to achieve them.

 Computers as Persuasive Media:

Conceptualizes persuasive technologies

supported by computers aimed at changing users’

behaviours and/or attitudes through hypothetical

experiences.

 Computers as Persuasive Social Actors:

Characterize technologies that use the computer

as a social actor whose goal is to change the

behaviour of its users by exploiting the social

dimension of technologies to influence users. This

is the kind of relationship that occurs when the

device plays the role of a teacher, coach, or expert

(Davis et al., 1998).

 Credibility and Computers: Conceptualizes

seven strategies according to the principle that

only credible computing products can influence

and convince users to change their opinions,

attitudes or behaviours.

 Credibility and the World Wide Web:

Conceptualizes the strategies defined for use on

websites whose objective is to convince users that

a specific website is the best way to access certain

information or certain content.

 Increasing Persuasion through Mobility and

Connectivity: Conceptualizes the strategies

defined for use on mobile devices that persuade

users in the appropriate time and place.

(Oinas-Kukkonen and Harjumaa, 2008) defined

persuasive system as “a computerized software or

information system designed to reinforce, change or

shape attitudes, behaviours or both without using

coercion or deception”. They proposed a framework

called Persuasive Systems Design (PSD) (Oinas-

Kukkonen and Harjumaa, 2009) for the development

and design of persuasive systems, composed into

three distinct phases: (1) understanding key issues

behind the persuasive system, (2) analysing the

persuasion context and (3) designing the system

qualities.

The first phase is concerned with understanding

the key issues behind the persuasive system, where

they define seven postulates that need to be addressed

on designing or evaluating a persuasive system, as

part of understanding the issues that underlie it.

The second phase of Persuasive Systems Design

involves understanding the context of persuasion,

which in itself consists of three aspects: intent, event,

and strategy.

 The Intent: Involves understanding (1) who is

persuader, and (2) what type of change is desired

(change of attitude or behaviour).

 The Event: Involves analysing (1) the use

context: what is happening around the behaviour

or the habit, (2) the user context: user’s goals and

(3) the technology context, which are the system

goals.

 The Strategy: The strategy must be clearly

defined, which consists of understanding: (1) the

Message that will be sent, and (2) the Route (how

it will be sent).

The third phase of persuasive system

development is the design of the system qualities.

Where they define four supports of persuasive

technologies that regroup 28 strategies, essential to

the establishment and maintenance of persuasion:

 Primary Task Support: Includes the design

principles that help users to carry out their main

task. Which are reduction, tunneling, tailoring,

personalization, self-monitoring, simulation, and

rehearsal.

 Dialogue Support: Includes the design principles

related to the establishment of communication

between human and machine (product). Which

are praise, rewards, reminders, suggestion,

similarity, liking, and social role.

 System Support Credibility: Includes the design

principles that describe how to create a more

credible and thus more persuasive system. Which

are social facilitation, social comparison,

normative influence, social learning, cooperation,

competition, and recognition

 Social Support: Includes the design principles

that leverages social influence to motivate the

users. Which are social facilitation, social

comparison, normative influence, social learning,

cooperation, competition, and recognition.

User Time Spent between Persuasiveness and Usability of Social Networking Mobile Applications: Patterns of Inﬂuence

3 RELATED WORKS

We classified existing related work into three

categories:

 The first category presents researches that

determine models of persuasion.

 The second groups presents the approaches that

determine the patterns of persuasion in online

social networks.

 The last groups presents researches that determine

the patterns of social influence, in social networks

and the persuasive systems in general.

The first category presents researches such as

those of (Fogg and Eckles, 2007). They analysed the

persuasiveness of over 50 “Web 2.0” services. Based

on this analysis, they identified a model applied by all

major social networks, called “Behaviour Chain

Model”. It outlines three Phases: “Discovery”,

“Superficial Involvement”, and “True Commitment”.

The first phase, “Discovery”, aims to preconize the

service provided by encouraging users to become

familiar with the site and to visit it. Once users

endorse to the service, they enter the second phase:

“Superficial Involvement”. During this phase, the site

highlights the possibilities of creating content and

exchanging information with other users in order to

motivate them to try the service. The third phase,

“True Commitment”, is characterized by three keys

target behaviours: (1) creating value and content, (2)

involving others, and (3) staying active and loyal. The

model gives important information about the stages

of behaviour change. However, it does not present the

influence strategies that support it.

The second category describes persuasion

patterns. (Iosub et al., 2009) have examined the social

application features of the three main Romanian

Social Networking Sites (SNSs), namely Facebook,

Hi5 and Neogen, on six dimensions corresponding to

the principles of social influence of (Cialdini, 1984) :

Reciprocation, Consistency, Social proof, Liking,

Authority, and Scarcity.

Their purpose was to determine how the

principles of social influence were integrated into the

design of SNS in order to achieve marketing goals, as

well as to determine which principles of social

influence are not currently used in the design of SNS,

and their application could potentially have a relevant

and value-adding effect on SNS competitiveness.

They analysed the three SNS from three

perspectives: social influence principles - marketing

goals, social influence principles - social networking

site and marketing goals - social networking site, and

they showed that social networking sites are mainly

organized according to social principles (reciprocity,

authority, social proof and taste).

In the same context (Adaji and Vassileva, 2016)

sought to identify the persuasion principles of Q&A

social networks. As a case study, they analysed the

StackOverflow network (social network where users

ask and answer questions related to IT). The authors

used the model PSD to analyse the social network

where they identified all persuasive principles present

in the PSD model, with the exception of just four:

tunnelling, rehearsal, reminders and similarity. The

authors conclude that each principle identified in the

social network StackOverflow seeks to promote the

engagement of users on the network.

However, their work was not comprehensive. The

first paper focalized on social design features under

relevant theoretical framework derived from social

psychology where the second paper focused just on

one type of social networks.

The last category presents researches that

emphasize only social influence and its patterns. For

instance, (Weiksner et al., 2008) studied Facebook as

a persuasive technology using a Grounded Theory

approach. They found and named six persuasion

patterns. They classified these patterns into two

categories: Native Patterns, relying on the

functionality provided by Facebook, and Adapted

Patterns, relying on functionality that was designed

for other contexts (e.g., desktop software, web sites,

video games, etc.).

The Native Patterns are composed of:

 Provoke and Retaliate: Allow a user to act

towards another user, by exploiting the principle

of reciprocity, the recipient of the gesture has the

social obligation to respond, according to the

norm of reciprocity.

 Expression Pattern: Allow users to create

artifacts to express opinions or affiliations, etc., by

exploiting the psychological principle of

ingratiation.

 Reveal and Compare: Allow a group of users to

act towards another group of users, by exploiting

the principles of reciprocity, cognitive

dissonance, the need to belong and social

validation.

 The Group Exchange: Allow a group of users to

create artifacts collectively following the

principles of impression management,

reciprocity, social validation, and context

variables such as feedback and applause.

The Adapted Patterns are composed of :

 Competition: Consists of enhancing competition

between users in order to motivate them to do

particular actions.

HUCAPP 2020 - 4th International Conference on Human Computer Interaction Theory and Applications

 Deception: Exploits the trust already acquired

from users to persuade them to achieve other

desired goals.

(Oduor et al., 2014) proposed software design

patterns for social influence. They aim to facilitate the

development of effective persuasion systems and to

make a linkage between the intent of designers and

users’ interactions. They proposed four patterns: social

learning and facilitation, competition, cooperation and

recognition. The patterns for social learning and

competition are based on sharing traces, whereas

cooperation and recognition rely on rich exchanges of

support between users.

Nevertheless, those works focused solely on the

social dimension and they did not take into

consideration all the other strategies and influence

processes present in the social network.

To the best of our knowledge, there is no study in

the literature that has extracted all the strategies of

influences into a set of well-defined patterns in an

inclusive way. An approach that presents, explains and

regroups the different strategies of influence used in

the different social networks, considering the different

influence dimensions (social, mental, etc.). In addition,

in our study, we present a direct relationship between

the identified patterns and the usability/persuasion

theory, and thus the criteria that have been

overexploited.

4 STUDY METHODOLOGY

At first we proposed five (05) patterns according to a

classification of usability criteria (the eight (08)

usability criteria for the design and evaluation of HCI

(Bastien and Scapin, 1993)) and persuasion strategies

(of the PSD model proposed by (Oinas-Kukkonen and

Harjumaa, 2009)) that we realized in advance defined

in table (Table. 1).

Table 1: Description of the five (05) proposed patterns.

Pattern

name

Definition Examples

Suggestion

Pattern

The fact of presenting the right information at the

right time gradually and in an allusive way,

compatible with the characteristics of the user, to his

needs, interests, and usage context, with the aim of

capturing his attention, in order to bring him to

engaging actions and to ensure his allegiance.

 Providing “Friends Suggestion”,

“Recommended Pages” and “Suggestions of

groups” functions on Facebook.

 Providing a set of related videos under the

video selected by the user on YouTube.

Reminder

Pattern

Any messages that interrupt users immediately,

leading them to respond to the delivered messages

asynchronously. It can take various forms, such as an

email, a social notification or an SMS.

 Facebook, YouTube, LinkedIn, Twitter or

Instagram notifications.

 WhatsApp, Messenger or Snapchat, messaging

system.

Social

influence

Pattern

Any form of influence that emphasizes the social

influence (comparison, competition, reciprocity, etc.)

in order to reinforce the pressure of other members on

the actions of an individual.

 Number of views for each video (on YouTube).

 Number of likes of each posts (on Facebook).

 Number of followers/tweets (On Twitter).

Interaction

Pattern

Any form of incitement or restriction that makes

targeted actions easy and simple while gradually

increasing the cost of the actions requested, or

deterring unwanted actions in order to make them

more difficult than they need to be.

Generally, social media interfaces facilitate tasks that

they want users to perform such as (inviting and

connecting to friends, registering in groups, adding

more pages), by making them simple actions.

 Example: (“Invite”, “Like”, “Share” and

“Register” buttons).

And complicate the unwanted actions that they do not

want the user to perform (disable notifications, delete

an account, etc.) for example:

 On Facebook, user cannot turn off all

notifications. He can just choose the mode and

the nature of the notifications he receives.

 Deleting an account on Snapchat requires

multiple steps as well.

Reward

Pattern

Any form of encouragement of user’s interaction

through forms of rewards that come to reward the user

in the success of his first interactions. It aims to

reinforce his initiatives and maintain his engagement.

 News feed with an infinite scroll (Facebook).

 Autoplay feature (YouTube) unending videos.

User Time Spent between Persuasiveness and Usability of Social Networking Mobile Applications: Patterns of Inﬂuence

Afterwards and in order to validate these patterns,

we adapted a hybrid card sorting method, implemented

on the web using the OptimalSort tool (Figure 2).

Figure 2: The OptimalSort Card Sorting Interface.

(http://www.optimalworkshop.com/).

We invited fifteen (15) eligible HCI experts to

participate in our study (5 men, 10 women), among

them, 9 are specialized in User

Experience/Ergonomics, 2 in Social

Computing/Human-Computer Interaction, 3 in

Multimodal interaction / Human-Computer

Interaction, and the remaining one in Human-

Computer Interaction and software engineering. Ten

(10) of them have more than ten (10) years of

experience in the field of HCI.

We believe that with the data of 15 experts we will

achieve the desired results since with 15 participants

the results obtained have a correlation of 90% as

affirmed by (Nielsen, 2004).

As cards, we choose the twenty-two (22) strategies

of persuasion of the PSD model proposed by (Oinas-

Kukkonen and Harjumaa, 2009), because it is the only

framework that exists, designed for the development

and design of persuasive systems. We choose also the

eight (08) usability criteria for the design and

evaluation of HCI proposed by (Bastien and Scapin,

1993). The thirty (30) criteria/strategies were presented

with definitions and examples.

Participants were invited to group the eight (08)

usability criteria and the twenty-two (22) strategies of

persuasion into categories. In addition to the

categorization that we proposed, we asked experts to

create their own categories if necessary, by assigning a

meaningful name to each category.

For analysing participants’ data, we adapted both

exploratory and statistical approaches. For the

exploratory one, we used Donna Spencer’s (Spencer,

2009) spreadsheets in order to understand each

expert’s categorizations, to create standardized

categories (which is an essential part of card sorting)

and to generate correlation matrix in order to

summarize the range of classification as well as to

provide information about each card’s placement and

the relationship between cards and participants. While

for the statistical approach, we used Hierarchical

Cluster Analysis (HCA) using complete linkage

method in order to highlight patterns in data.

5 RESULTS AND ANALYSIS

5.1 Standardizing Categories

At first, we inserted the data of each participant

extracted, in Donna Spencer’s (Spencer, 2009)

spreadsheet (Figure 3. Shows the generated matrix).

Each row represents a card and each column a

participant. The row and column intersection has the

label for each category. Then, we began standardizing

the labels of categories created by experts in order to

create a set of consistent categories’ labels.

Figure 3: Worksheet with all the raw data.

The experts validated our suggested categories.

They categorized the criteria / strategies using only our

five proposed categories, with the exception of three

experts, who proposed four additional ones. Thus, we

intersected our categories and the additional ones

proposed by experts based on similar words

(terminology of labels), similar ideas and categories

that contains similar cards.

For example, a participant created a category

named “Restriction” that includes the cards (Explicit

Control and Tunneling). However, we found that it was

a subcategory of the “Interaction” category (which

covers the “restriction” and the “incitement”). Another

example, two participants created two categories

named “Customization or Personalization” which

means: “The content customized to the needs of the

HUCAPP 2020 - 4th International Conference on Human Computer Interaction Theory and Applications

user” and “Data Influence” which means: “the data that

influences the user in order to attract him”, including

cards (Adaptability, Compatibility, Personalization,

Similarity, Suggestion and Tailoring). However, these

two categories means the same as the “Suggestion”

category, which means, “Offering fitting suggestions

to the user in order to attract and incite him”.

5.2 Cards Placement

After the standardization process, we had an average of

five (05) categories. Table. 2 recaps the standardized

categories, the percentage of each cards’ placement

and the content of each category as well as an insight

of the relationship between cards, categories and

participants. According to this table (correlation matrix

generated by (Spencer, 2009)), we notice that all

categories have at least one card

with high agreement

(75% or more of participants agreed to the placement

of the cards), and 60% cards belong to categories with

medium agreement (50% or more of participants

agreed to the placement of the cards).

Table 2: The correlation matrix created for the hybrid card-sort (as generated by (Spencer, 2009)).

Card name

Interacti

Remind

Rewar

Social

influen

Suggesti

Categori

es for

this card

Categori

es with

high

agreeme

Categori

es with

medium

agreeme

Categori

es with

low

agreeme

Adaptability 73% 0% 7% 0% 20% 3 0 1 2

Compatibility 33% 0% 0% 20% 47% 3 0 2 1

Competition 13% 0% 7% 80% 0% 3 1 0 2

Consistency 80% 0% 0% 7% 13% 3 1 0 2

Cooperation 13% 0% 13% 67% 7% 4 0 1 3

Error Management 47% 33% 7% 0% 13% 4 0 2 2

Expertise 0% 0% 20% 60% 20% 3 0 1 2

Explicit Control 80% 0% 7% 7% 7% 4 1 0 3

Guidance 60% 0% 0% 0% 40% 2 0 2 0

Liking 47% 0% 7% 40% 7% 4 0 2 2

Normative influence 7% 0% 7% 80% 7% 4 1 0 3

Personalization 40% 0% 0% 0% 60% 2 0 2 0

Praise 0% 7% 73% 20% 0% 3 0 1 2

Recognition 0% 13% 47% 33% 7% 4 0 2 2

Reduction 60% 0% 7% 7% 27% 4 0 2 2

Reminders 7% 93% 0% 0% 0% 2 1 0 1

Rewards 0% 7% 80% 13% 0% 3 1 0 2

Self-monitoring 27% 40% 7% 13% 13% 5 0 2 3

Significance of Codes 73% 13% 7% 0% 7% 4 0 1 3

Similarity 13% 13% 13% 20% 40% 5 0 1 4

Social comparison 0% 0% 7% 93% 0% 2 1 0 1

Social facilitation 0% 0% 0% 100% 0% 1 1 0 0

Social learning 0% 0% 0% 100% 0% 1 1 0 0

Suggestion 7% 0% 0% 13% 80% 3 1 0 2

Surface credibility 33% 0% 7% 33% 20% 4 0 2 2

Tailoring 20% 7% 0% 0% 73% 3 0 1 2

Trustworthiness 27% 0% 7% 40% 20% 4 0 2 2

Tunneling 53% 0% 7% 0% 40% 3 0 2 1

Verifiability 40% 0% 13% 27% 13% 4 0 2 2

Workload 87% 7% 0% 0% 7% 3 1 0 2

Cards in this category 23 10 20 21 23

Cards with high agreement

(>75%) 3 1 1 5 1

Cards with medium

agreement 13 2 2 7 7

Cards with low agreement

(<25%) 7 7 17 9 15

User Time Spent between Persuasiveness and Usability of Social Networking Mobile Applications: Patterns of Inﬂuence

 Cards in this Category: Refer to the number of

individual cards placed in this category.

 Cards with High Agreement: Refer to the

number of individual cards with a correlation of

75% or over (75% of participants or more used

this category for the card).

 Cards with Medium Agreement: Refer to the

number of individual cards with a correlation of

25%–50%.

 Cards with Low Agreement: Refer to the

number of individual cards with a correlation of

25% or less.

 Categories for this Card: Refer to the number of

categories used for this card.

 Categories with High Agreement: Refer to the

number of categories with a correlation of 75% or

over (75% of participants or more put this card in

a consistent category).

 Categories with Medium Agreement: Refer to

the number of categories with a correlation of

25%–50%.

 Categories with Low Agreement: Refer to the

number of categories with a correlation of 25% or

less (participants or fewer used any consistent

category).

Next, we applied an HCA to the global data matrix

of the 15 experts, we were able to identify and label

the clusters on the dendrogram generated by the

HCA. Figure 4 illustrates these results.

5.3 Patterns Extraction

According to the dendrogram (Figure 4), five

consistent patterns are distinguished from the

classification and are cited below:

 Suggestion Pattern: This pattern contains four

persuasive strategies from PSD Model:

Suggestion, Personalization, Tailoring and

Similarity and one usability criteria from (Bastien

and Scapin, 1993) ergonomic criteria:

Compatibility.

 Reminder Pattern: This pattern contains one

persuasive strategies from PSD Model:

Reminders.

 Social influence Pattern: This pattern contains

seven persuasive strategies from PSD Model:

Social comparison, Social facilitation, Social

learning, Expertise, Normative influence,

Competition and Cooperation.

 Interaction Pattern: This pattern contains seven

persuasive strategies from PSD Model:

Reduction, Tunneling and to a lesser extent

Verifiability, Liking, Surface credibility and

Trustworthiness, and six usability criteria from

(Bastien and Scapin, 1993) ergonomic criteria:

Workload, Adaptability, Consistency, Explicit

Control, Significance of Codes and Guidance.

 Reward Pattern: This pattern contains three

persuasive strategies obtained PSD Model:

Rewards, Recognition and Praise.

Figure 4: Dendrogram generated by the HCA of the hybrid

card-sort.

5.4 Patterns’ Relationships and

Proximity

From Figure 4 we notice that there is a close

relationship between patterns. The pattern

“Interaction” is close to the pattern “Suggestion”, and

both are in turn related to the pattern “Reminders”.

This explains the design of social media that aims to

attract the attention of the user and arouse his interest

to undertake an engaging action, favourable to the

media. On the other hand, the pattern “Social

Influence” is related to the pattern “Rewards”,

HUCAPP 2020 - 4th International Conference on Human Computer Interaction Theory and Applications

applied to complete the total commitment of the user

in order to control him.

6 DISCUSSION

In this paper, we identified five influence patterns that

bring together all social media strategies having a

significant impact on user time by applying hybrid

card sorting method. We consider that those patterns

are usually used together gradually since they support

each other (as shown in Figure 5). In fact, the process

of capturing the user is a temporal process that

includes a start, followed by modifications of the user

behaviour towards the media, leading to the objective

sought (maximizing the UTS). Moreover, to bring the

users to this objective, a social network must bring the

user gradually over time, with a succession of phases,

to capture their attention and especially to maintain

their interest and their motivation.

First, the social media must solicit the user to take

the first steps over the social media by offering him

suggestions (proposing close people, adhere to

groups, etc.), by exerting (social) pressure using the

mechanism of normative influence or social

comparison for example, or by reminding him of

actions that he can perform (using instant

notifications for example).

Moreover, to have the prescribed goal and the

behaviour expected from the user, the social network

must prompt him to take action where it is necessary

using simple interactions (for example, buttons that

say click here or subscribe now). On the other hand,

the social network can complicate the unwanted

actions desired by users (for example, disable the user

account, or disable notifications).

Finally, the social media must reward the user for

the success of his first interactions, in order to

reinforce his initiatives and behaviours. Nonetheless,

this process of patterns sequence is our own

conclusion that may need some further research.

All these patterns are useful regarding usability

and persuasion. They guide users, give them some

control, present urgent messages, suggest desired

videos, etc. Nevertheless, these results highlight

important negative points (unlimited suggestions list,

unlimited notifications at any time in the day even if

there is no emergencies, etc.) encouraging users to

use applications more frequently and for greater

duration than they otherwise would.

In addition, patterns such as “Interaction Pattern”

impose an important issue: complicating some

interactions such as reducing notifications or

disabling accounts is not recommended according to

usability criteria. According to (Christian Bastien and

Dominique Scapin, 1993), the usability criterion

“Minimal Actions” consists of reducing the number

of actions necessary to accomplish a goal or a task

(“limiting as much as possible the steps users must go

through”). Thus, complicating some actions calls into

question the interaction pattern and its use in

ergonomics (especially in persuasion).

Figure 5: Relationships between patterns.

To summarize, exploiting all these patterns in

usability and persuasion must be questioned not only

because they have a big effect on the user's time, but

also because their misapplication can conflict with the

ergonomics theory and overexploit its criteria.

7 CONCLUSION

In our paper (Bedjaoui et al., 2018), we showed that

there are some gaps in the HCI ergonomics theory

that gave rise to addictive interfaces. These gaps lie

in the fact that designers and/or developers have to

use their judgments to implement the different

usability and persuasion criteria, principles and rules.

In addition, these criteria have no application

borderline. Thus, designers or developers can apply

them according to their needs or interests without

necessarily considering the users objectives (how

they want to spend their time), their relationships

(interpersonal and social choices), their mental and

emotional limits (vulnerabilities, tiredness and ways

in which their minds form habits), etc.

In this paper, we identified five influences

patterns using hybrid card sorting method. Those

patterns have a significant impact on user time spent

on the media and their misapplication can conflict

with the ergonomics theory.

User Time Spent between Persuasiveness and Usability of Social Networking Mobile Applications: Patterns of Inﬂuence

Now, we are working on an approach that frame

these influence patterns. We are defining their

application borderlines. For a particular pattern

applied to support particular criteria, we are

identifying rules that make its use more efficient

(from user time viewpoint), keep its advantages and

do not affect the supported criteria. Thus, while

applying a pattern, designers or developers will be

aware of the usage limits and thereby respect them to

benefit from the mechanism advantages while

respecting the users’ time.

We believe that such an approach will allow to

design and build products that motivate the active

participation of the user and respect her/his time.

REFERENCES

Adaji, I., & Vassileva, J. (2016, April). Persuasive patterns

in Q&A social networks. In International Conference

on Persuasive Technology (pp. 189-196). Springer,

Cham.

Bastien, J. C., & Scapin, D. L. (1993). Ergonomic criteria

for the evaluation of human-computer

interfaces (Doctoral dissertation, Inria).

Bedjaoui, M., Elouali, N., & Benslimane, S. M. (2018,

November). User Time Spent Between Persuasiveness

and Usability of Social Networking Mobile

Applications: A Case Study of Facebook and YouTube.

In Proceedings of the 16th International Conference on

Advances in Mobile Computing and Multimedia (pp.

15-24). ACM.

Cialdini, R. B. (1984). The psychology of persuasion. New

York: Quill William Morrow.

Davis, J. W., & Bobick, A. F. (1998, November). Virtual

PAT: a virtual personal aerobics trainer. In Workshop

on Perceptual User Interfaces (pp. 13-18).

Fogg, B. J. (2002). Persuasive technology: using computers

to change what we think and

do. Ubiquity, 2002(December), 5.

Fogg, B. J., & Eckles, D. (2007, April). The behavior chain

for online participation: how successful web services

structure persuasion. In International Conference on

Persuasive Technology (pp. 199-209). Springer, Berlin,

Heidelberg.

International Organization for Standardization. (1998). ISO

9241-11: Ergonomic Requirements for Office Work

with Visual Display Terminals (VDTs): Part 11:

Guidance on Usability.

Iosub, D., Andrei, A. G., & Iacob, A. (2009). Patterns of

social influence in social networking sites–a design

perspective.

Kuss, D. J., & Griffiths, M. D. (2011). Online social

networking and addiction—a review of the

psychological literature. International journal of

environmental research and public health, 8(9), 3528-

3552.

Némery, A., & Brangier, E. (2014). Set of guidelines for

persuasive interfaces: organization and validation of the

criteria. Journal of Usability Studies, 9(3), 105-128.

Nielsen, J. (1994). Usability engineering. Elsevier.

Nielsen, Jakob. Card Sorting: How Many Users to Test.

https://www.nngroup.com/articles/card-sorting-how-

many-users-to-test/ (2004)

Oduor, M., Alahäivälä, T., & Oinas-Kukkonen, H. (2014).

Persuasive software design patterns for social

influence. Personal and ubiquitous computing, 18(7),

1689-1704.

Oinas-Kukkonen, H., & Harjumaa, M. (2008, February).

Towards deeper understanding of persuasion in

software and information systems. In First international

conference on advances in computer-human

interaction (pp. 200-205). IEEE.

Oinas-Kukkonen, H., & Harjumaa, M. (2009). Persuasive

systems design: Key issues, process model, and system

features. Communications of the Association for

Information Systems, 24(1), 28.

Shneiderman, B., Plaisant, C., Cohen, M., Jacobs, S.,

Elmqvist, N., & Diakopoulos, N. (2016). Designing the

user interface: strategies for effective human-computer

interaction. Pearson.

Spencer Donna. Card Sort Analysis Spreadsheet Template.

Retrieved November 2016, from

http://rosenfeldmedia.com/books/card-sorting/ (2009)

Weiksner, G. M., Fogg, B. J., & Liu, X. (2008, June). Six

patterns for persuasion in online social networks.

In International conference on persuasive

technology (pp. 151-163). Springer, Berlin,

Heidelberg.

HUCAPP 2020 - 4th International Conference on Human Computer Interaction Theory and Applications