1.5.3
Personal and Identifiable Data
In this research work, we have collected personal information and identifiable data in different stages, as part of experiments and pilot prototype deployments. Data has been collected implicitly from system log files, automatically via auto- matic capture equipment (e.g., a lifelogging camera) as well as, explicitly using physiological measuring equipment, questionnaires, interviews, field notes, and audio and video recordings. Developed interventions and prototypes have also collected data for fulfilling their task (e.g., keeping record of a participant’s daily activities, cognitive and physiological states).
The interactions of subjects with a prototype as well as, their demographic background information have been collected in order to evaluate an interven- tion’s memory augmentation performance and its usability. The identity of the subject, typically contact information such as name and e-mail addresses, have been stored separately from data collected during individual user studies. Sub- jects have been known within and across studies only by a pseudonymous iden- tifier. In addition, we have also collected various background demographics in order to better classify and understand experimental results. This data is not personally identifiable. The exact demographic information collected had been determined at the time of experimental design, but typically included age, gen- der, education, and occupation. The exact data collected has always been de- tailed in the ethical worksheet drafted before and for each experiment. All data collections have been conducted according to the corresponding national and in- stitutional regulations, including notifying the appropriate national institutional organizations when required, and always after the approval of the assigned eth- ical board.
The personal information of study subjects (i.e., their identity information) have been stored in a centralized fashion and in encrypted form. Any user pro- file created during the interaction with a memory augmentation intervention has been encrypted within the system, with access keys only available to subjects and the involved researchers. Throughout the duration of this work, standard en- cryption techniques have been used for storing data collected locally or remotely during experiments.
1.6
Thesis Outline
This thesis is comprised of 12 chapters, organized in 3 parts, followed by a bibli- ography section. In Part I – Background, we present the motivation, theoretical
14 1.6 Thesis Outline
underpinning and methodological background of this work. Part II – Studies, en- compasses the research conducted for augmenting episodic and semantic mem- ory recall. In the last part (Part III – Synthesis), we draw on obtained knowledge from field deployments and literature review for answering our research ques- tions, we describe our contributions, and we summarize the outcome and the future of this work. Below, we provide a brief description for each chapter in each part of this thesis:
1.6.1
Part I – Background
• Chapter 1 – Introduction. The first chapter describes our vision and moti- vation for augmenting memory recall with ubiquitous technologies, defines the research questions this thesis answers, outlines the research context based on which this thesis is written, and summarizes the key challenges we faced throughout this endeavour, including overall contribution. • Chapter 2 – Foundations. In the second chapter, we provide a synopsis
on the structure of human memory and the processes that pertain to it. Then, we showcase how modern (technological) trends and practices such as lifelogging and the quantified-self movement could support memory re- call through the appropriate actuation of captured contextual information. We also include insights gained from an extensive literature review on Lifel- ogging User Interfaces (LUIs), eliciting design principles and guidelines to which LUIs adhere and the goals they fulfil.
• Chapter 3 – Augmenting and Measuring Memory Recall. In this chap- ter we describe our methodology and in particular how we transferred an established psychological method such as "cued recall", from the lab into the wild. We present the apparatus we used, and we outline the meth- ods we applied for collecting contextual information, generating memory cues, delivering memory cues, and finally evaluating the effectiveness of our memory-augmenting interventions.
1.6.2
Part II – Studies
• Chapter 4 – Event-Driven Image Capture for Augmenting Episodic Mem- ory Recall. Here we describe a first user study that investigated the poten- tial of event-driven captured self-face pictures (event-driven selfies), taken at a common mobile interaction moment (i.e., screen unlock) via a dedi- cated mobile application that utilizes the front–facing camera of a typical smartphone, to effectively trigger episodic memory recall. The potential
15 1.6 Thesis Outline
of event-driven selfies in supporting episodic memory recall was tested in several stages after capture.
• Chapter 5 – Augmenting Memory Recall for UX Evaluation. In this chap- ter, we demonstrate how event-driven captured selfies in the form of mem- ory cues can benefit the field of retrospective UX evaluation for mobile applications. We present a user study with regular commuters predicting, experiencing and recalling negative emotions (i.e., anger and frustration) during the experience of daily commute. Particularly, we were able to un- veil plausible effects on how commuters recall a negative driving experi- ence with potential significant implications for in-car UX design. Leverag- ing on the the potential of selfies to trigger memory recall, we showcase an application prototype that features a combination of memory cues (i.e., pic- tures, video of facial expressions, location, etc.) for safely eliciting driver’s affective state at a later stage. We believe our prototype can benefit urban planners and in-car UX designers by retrospectively measuring in-car UX levels.
• Chapter 6 – Capture Modality Effect on Memory Recall. Here, we present an extensive user study that compared three different picture capture modal- ities (i.e., unlimited, limited, and automatic) in their effect on memory re- call, with and without the support of the captured pictures as memory cues. The study explored novel design dimensions for future pervasive memory augmentation systems, by artificially limiting the amount of pictures one can take via a dedicated mobile application, while comparing it with preva- lent mobile picture capture, and a popular wearable lifelogging camera. • Chapter 7 – Augmenting Memory Recall for Work Meetings. In this
chapter, we report a user study that inquired into the effectiveness of recall- augmenting interventions in the workplace context by recruiting partici- pants that meet regularly in a "supervisor-supervisee" fashion. By utilizing a combination of episodic and semantic memory cues, we were able to es- tablish a quantifiable baseline of memory augmentation in work meetings that serves as reference for future research on pervasive memory augmen- tation systems.
• Chapter 8 – Physiological Responses in VR Experience Recall. Here, we present two studies that highlight the potential of physiological responses in driving the selection of memory cues that could trigger episodic memo- ries at later stages. We investigate the interplay of physiological responses