Journal Article: Electronic Partners for Improved Support of User Values

Currently writing my second journal article. This article highlights that location sharing apps, especially in the family domain, need to provide a better support for a certain set of domain-related user values (e.g. family security, friendship, social recognition, and independence). The article relies primarily on the series of  user studies at the day care centers, which I wrote about earlier here, here, and here. It picks up from the results of first paper which evaluated a social commitment model’s usability and usefulness, and goes on to show that location sharing apps augmented with such a model, i.e. “electronic partners”, can provide enhanced support for user values.

App user test @Zo Kinderopvang, Den Haag.

The third and last of a series of user studies for our app, was conducted at Zo Kinderopvang daycare center, Den Haag. Thirteen children participated in the first testing sessions, but unfortunately two of them dropped for the second session. The procedure followed was similar to the one here.

Brainstorming on some of the privacy models of some location sharing apps

In the effort of trying to create a spectrum of autonomy/accountability for location sharing apps that can be used in the family domain, and place our app on that spectrum as well, I’ve come up with this sharing/receiving analysis for some of the apps that appear to be on the “extreme” sides of the spectrum:


mSpy is an app A can install on B’s phone to send all types of data (including location updates) to A.  B cannot disable the app and may not even be aware it is installed.


Life360 allows users to create and join (or refuse invitations to join) “circles”.

A user can define with which circle their location updates (x minute interval updates of their GPS position) are shared (e.g. yes with family, but not with colleagues) at any time.

(Not sure if relevant in this paper’s context, but 1. The above implies a traceable history, limited to one day backwards for non-premium users, and 2. Users can create locations to get notified when members of certain circles enter those places).

Users can check-in, which means they share their GPS position with all their circles. This cannot be enabled/disabled for different circles (considered to be a voluntary declaration of their whereabouts).

There are other features such as the “panic button” which sends a panic alert (plus text messages and emails) with your location to every person in your circles.

High autonomy for the child, because Life360 cannot oblige them to share any info they do not want to share or join any circles they do not want to join. But it suffers from the same disability as our non-commitment ePartner (e.g. you can’t promote independence without demoting family security).

Swarm (4sq)

  • With Swarm we can make the distinction between “friends”, “venue managers” and “public”.
  • Users have the ability to request friendship and accept (or refuse) friendship requests from others. Users have the following for privacy options:
  • They can select where their check-ins (place-defined check-ins, like cafés or bars) are shared with venue managers (if the check-in happens in their venue) and/or the public (if they are currently in that venue). Check-ins are always shared with friends.
  • Check-ins are not forced to match GPS location (unlike the above two apps)
  • They can enable a “neighborhood locator” which uses GPS-location to reveal location on the neighborhood level to friends at all time (e.g. Manhattan, Scheveningen, etc.) if no recent check-in has been made for a period of time. This is correct info and cannot be manipulated like check-ins.
  • Swarm can post to third party apps (e.g. Facebook and Twitter) where that app’s privacy settings override that of Swarm.

Friends can include you in their check-in, apparently you cannot prevent them from doing so, even if that was an incorrect check-in. Apparently you can only decide whether that check-in will automatically appear on their Facebook or Twitter feed with your name included. (I need to check this item more).

App user test @De Lange Keizer, Delft

Description of the first of a series of user studies for our app, this one conducted at De Lange Keizer daycare center, Delft. Eight children participated in the two testing sessions.

This user study was the first field test of our implemented app prototype. The app is a location sharing device for families with children in primary school (between 6-12 years of age), which allows users to create locations, check-in in these locations, share and receive check-ins from others as they choose, and create normative-based social commitments with other users regarding sharing/receiving check-ins. We believe that this additional capability (social commitments) will provide a better support for user values for children, namely social recognition, friendship, independence, and freedom (amongst other hypotheses).

We have built two editions of the app. Edition 1 comes without social commitments (a.k.a. afspraken) and edition 2 comes with social commitments. These hypotheses/research questions we are attempting to answer in this user study are:

  1. Within the domain of family life, e2 will provide better overall value support for children than e1.
  2. Within the domain of family life, e2 will be perceived more as social actor than e1.
  3. Within the domain of family life, e2 will be perceived more as useful/usable than e1.
  4. (Research question) what subset of the possible social commitments will our users use, and for what purposes/values?
  5. Children will expect that the presence of an app in their life will increase support to their values.
    1. Sub hypothesis: the children will expect that the presence of an app in their life will increase their independence (or personal freedom).
    2. Sub hypothesis: the children will expect that the presence of an app in their life will increase have a positive effect on their friendships (or their social recognition).
  6. Children will expect that the presence of an app in their life will increase support to their activities.
  7. Children will expect that the presence of an app in their life will remove some of the limitations they have.


The user study took place at a local day care center (Dutch: buitenschoolse opvang) in Delft. This day care center is a place where children are brought after school is over, until their parents pick them up at the end of their workday. It contains 9 different activity rooms where children can freely participate in activities, such as a painting room, music room, playground, lunchroom, etc. The day care center officials have assisted us finding children who wanted to participate in the user study.

Due to the number of participants (8 children, aged between 7 and 10 years) we opted for a within subject, counterbalancing design, meaning that children were randomly split in two groups of four, each group testing a different edition of the app in each of the two sessions.

Introductory session

We held an introductory session one week before the user study sessions, where the researchers met the participants, introduced the app using a tutorial video, and answered their questions about the app and the procedure of the user study.

User study sessions

The two sessions were identical, and they were one week apart. Below is a description of what happens in each session.

Children had to perform “missions” using the app. A mission is either:

  • Instructional, directly pointing to a functionality in the app to help the user get acquainted with the app more. Examples: checking-in in a certain location, adding other users to their friend or family list, and creating a social commitment.
  • A story-lined, simulation of a real life situation, that requires the user to figure out what needs to be done with the app without direct instructions. Examples: go to the lunchroom. Did you find any of your friends there? How can you ensure that one of your friends will inform you when they enter this location? Check who’s in the painting room. Anyone from your team? If yes, try to tell all your friends that you are both at that location. Can you see where others on your team are? Try to go out and find one of them. Hint: check your event list.

There were 37 mission cards in total (17 instructional missions, 20 real-life situation simulations, see example in photo below).  The missions had an even distribution of “envisioned” solutions in terms of app functionalities and the possible types of commitments that need to be created.

Children would pick random missions from the pile, read the text, (attempt to) perform them, and then bring the mission back when they would like (whether finished successful or not) to the pile to pick another one, and so on. The user study stopped when one hour had passed in each session.

During which the first 10 minutes only instructional missions were available to be picked from the pile (in order to build some app knowledge). Afterwards, we added the rest of the missions to the pile.

Collected data

  1. Value measuring questionnaire: answered by children at the end of sessions 2 and 3. Contains questions to measure fulfillment of the values Friendship, Social recognition, Independence, Freedom. (Hypotheses 1,5,6,7). Questions in this questionnaire were framed into a hypothetical future tense: e.g. “if I had this app in my life, it would <less easy………..nothing will change……..much easier> to find where my friends are. Before they answered the questionnaire, children watched a short video explaining how the questionnaire should be answered, with a period for questions and answers.
  2. Questionnaire for usability and social actorship, also answered by children at the end of sessions 2 and 3. Contained questions for usability, liking, dominance, trust, and intimacy. (Hypotheses 2,3).
  3. Every time a child picked a mission from the pile, the id numbers of the child, mission, and timestamp were recorded. (Research question 4)
  4. Behavioral sampling: every 10 minutes, one of the researchers recorded every child’s emotion/body language according to a predetermined coding schema (derived from Markopolous, Evaluating Children’s interactive Products, pp. 182, with positive codes created by reversing the negative ones already existing in the text), their location, and their engagement with others, i.e. whether in groups or alone. (Hypotheses 1,5,6,7).
  5. All app interactions with any user were recorded and time-stamped on the server, e.g. every check-in, every time a user added another user to a list, every time a commitment was created, accepted, rejected, etc. (Research question 4).


Video to help explain to children how to answer questionnaire

Just created a new video (again, using iMovie and the GoAnimate website), to try and bridge the gap for children between the app they are going to be testing, and the futuristic questions they have to answer in the questionnaire. I have to admit it is difficult to imagine how a 7-10 year old child can answer a “fill in the blank” questionnaire with a continuous slider like this shown in the video. Hopefully this tutorial will ease the transition!

Demo (and paper) accepted to UMAP 2014!

My demo (and included paper) “A Value-Sensitive Mobile Social Application for Families and Children” accepted for the 2014 UMAP (User Modeling, Adaptation, and Personalization) conference, which will be take place in the the Danish city of Aalborg in June.

I will be showing my app on the third day of the conference during the demos event, and conference participants will be able to try the fully functioning version. Looking forward for the experience!

Pilot testing the app (2)

Today I went back to Hendrik-Ido-Ambacht. I met with the same pilot test group at a community center. The children have used the app quite extensively, but the parents/adults, not that much. One of the explanations was given by the parents was that they already use other means of communication, like WhatsApp, SMS’s, Facebook, etc. One other important aspect, well, this group was quite familiar with each other, but in real future tests, Parents shouldn’t be allowed to make agreements with other people’s children, and vice-versa– it would be considered somewhat strange. Plenty of other feedback came out of this short, 40-minute focus group: one of the parents found that the mandatory buzzer everytime they received a check-in or an agreement request was not needed, it was a little too much. She also suggested using the Android notification center, which I have gone to great lengths to bypass for this app, because I believed it was too hidden/small fonted (especially on this screen)/difficult for children between 7-11 to understand– based on standard design-for-children advice. Apparently for the two participating children in this pilot, this was not the case. We do really underestimate what children are capable of, tech-wise I’d say.

Plenty of lessons to learn here when designing for children. First of all, plenty of literature and books on the subject are North American– where the ages at which children learn reading, writing, and the activities they do, they all differ from European school systems. Also, the use of the word children itself could mean completely different groups, as most of the advice you get from such books is aimed towards children 3-5 years of age, which still can’t read for example. If designing for an older group, one shouldn’t worry about incorporating more of the “advanced” interface elements that adults usually use. Cognitive skills improve really rapidly at this point too– even half a year or one year of growth would make design decisions somewhat obsolete.