For german version click here
Smartphones are becoming increasingly important in our lives and accompany us almost everywhere. It has thus become possible to ask the device for directions, check email, spend time playing video games or simply surf the Internet.
In the following study we would like to scientifically assess, how often people are actually using their mobile phones. When do we go online for the first time a day, when do we first check for new email and when do we receive the first call? Additionally, it is the aim of the study to find out how differences in smartphone usage can be explained. In this context we would like to take into account classic self-report inventories from personality psychology (e.g. recording the self-perceived anxiety or sociability). More specifically, you have the chance to fill in a small number of items on your personality and get immediate feedback on your personality profile.
Because your smartphone contains very private information, we would like to explain, what information the software stores and which it does not. It does not record any personal content from emails, SMS or chat messengers (e.g. WhatsApp). Of course it does not record, what internet addresses you visit or with whom you are talking over the phone. We are interested in quantitative variables - how long are you online per day, when do you look on the phone for the last time in the evening and how big is your social network (how many different phone numbers are there, but not what phone numbers are in your phone book)? Subsequently, we elucidate what information we record, what the study’s design looks like and how we secure your data.
In the following we explain what information we record how the study design looks and how we protect your data.
Following the installation of an app developed by us (in the department of Computer Science) we will record your mobile phone behavior. Therein, we examine:
- Your interaction with your mobile phone and the apps installed on it (when you turn it on, when you use which app)
- Your communication patterns in rudimentary fashion, concretely how much you converse with which (anonymized) phone number and to what extent you communicate per SMS or messenger.
Here we record, ordered by aspects, the following variables:
- Your interaction with the mobile phone through so-called app sessions: We define an app session as the range of time that you interact with a certain app installed on your phone. Concretely this means that an app (i) runs in the foreground and (ii) you as a user interact actively with the phone. An app session ends when you (i) switch to another app (ii) quit the app or (iii) stop using the phone.
- We further observe how often you flick your phone’s screen on without actually unlocking it and how much time you spend in the menu navigation of your phone. Additionally we record all apps that are installed on your phone, independently of whether you actually ever use them or not. We do NOT record, what you do in the apps, i.e. what contents you read or write.
Communication patterns: We observe incoming and outgoing phone calls as well as SMS and chat messages. With this information we assess you according to the size of your social network. We merely investigate with how many contacts you are in regular contact and who initiates the call or the SMS. We assess NO information about the content of the calls or SMS. The software does not record text or sound. Additional communication channels (WhatsApp, Skype, Facebook) are not recorded (e.g. again we record if you use WhatsApp but not what you are writing or receiving via this channel). Phone numbers are only recorded in anonymized form.
The method presented in the following ensures that the original phone numbers can not be reconstructed from the sent data: The software already encrypts the contacted phone numbers on the phone by means of a so-called cryptographic hash function. From each phone number this generates a definite code, henceforth called anonymized phone number. It is proven mathematically to reconstruct the original phone number from this. The phone internally uses a key that it creates automatically upon the initial installation. The key is saved only on the phone and is deleted for all times upon the uninstallation. A reconstruction of the original phone numbers is therefore impossible. Similar to app sessions, we record for SMS and phone calls data points with timestamps. Namely for:
- – Phone calls: [anonymized phone number, start time, end time]
- – Incoming SMS and chat messages: [anonymized phone number, number of characters, time of sending]
- – Outgoing SMS and chat messages: [anonymized phone number, number of characters, time of reception, time that the message was read]
- – Movement patterns: We are here interested in the dynamics of the movement profile (but not where subjects are currently located), i.e. how much change in movement can be recorded with the help of the GPS signal. There is more information regarding this topic later under geographical position.
Additionally we extract all phone numbers saved in the address book and communicate these (in anonymized form, as mentioned above) to the server in the department of Jun. Prof. Markowetz (Computer Science, University of Bonn: also see further down)
Geographical position: Herein we record movement patterns. The positioning is performed with the GPS sensor built into the phone. Should this not work or if there is currently no signal (as is the case inside many buildings) the phone tries to estimate the position with the help of nearby cell towers and WIFI base stations. Both methods are supported directly by the phone’s operating system. The precision of the localization ranges from 10 to 100 meters. Concretely, a localisation is performed every 20 minutes and a data tuple (data entry) of the following form is built: [time stamp, longitude, latitude, precision]. The first three fields are self-explanatory, the last describes how exactly the phone thinks the effected measurement is. The description of the data tuple at this point does not contradict the fact, that we are only interested in your movement patterns regarding dynamics (see d). The data tuple is not evaluated to learn about your actual position but to create various inter-distance-measures, for example the diameter or the distance covered in one day.
Mental states, life satisfaction, and personality: the software asks for your mood two times per day (in the morning and the evening). To this end, you are also able to answer questions on life satisfaction or personality.
The data is sent from the phone to the servers in the division of Jun.-Prof. Alexander Markowetz, where it is stored and evaluated. Only select members of Markowetz’s and Montag’s working groups. Data will not be passed on to external people. The technical data protection (data security) is guaranteed at all times.
Anonymity of the data
Data is assessed in anonymous fashion. This is technically true only if the participant’s email address does not reveal any information about the name (e.g. email@example.com vs. firstname.lastname@example.org). We inform you herein that you have the possibility to register with such an anonymous email address. In the case that you already have such an anonymous email address it may still be a good idea to create another one specifically for this study. In any case we perform the statistical assessment of your data without regression to your name. Additionally, we inform you, that we are not interested in the results of individual participants. This means that your data is always only assessed in the context of the entirety of the recorded data. Generally the deletion of the data happens after ten years automatically, except if you wish an early deletion. This is possible at any point through the app. Your data will not be passed on to third parties but only analyzed by members of the department of the psychological institute of the University of Bonn or University of Ulm (Markowetz/Montag’s groups). Further information about the possible deletion of your data can be found under “voluntary participation”.
Nonhazardousness of the app
The app developed by us has been extensively tested and does not cause any damage on your mobile phone. This is true also for the general experience with daily downloaded apps: Actually, millions of users install apps, without any reports of devices being damaged by some app. An unintended damaging can be empirically precluded.
Voluntary participation in the examination
The participation in this examination is voluntary and can be ended at any point in time without giving reasons. That means that you can uninstall the app at any point in time. Canceling the study is related with any negative consequences for you. Another button in the app allows you to delete all the data recorded by us.
By clicking this button I allow the recording, storage and use of my mobile phone data to scientific ends. I have read the proband information and have no further questions.