literature for paper one! plus other notes - 27

Analysis of literature on mobile apps found on business and marketing journals.
Arora 2017 JM***: prices and revenue model and how this influences the demand of an app. This element could be used for the second research question. (as reminder: the 2nd rq requires the introduction of a paragraph on the introduction, plus two paragraphs on the theoretical frame about what is the best selection of a revenue model.
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Note concerning econometric model in business:
The selection of a model is not a random variable. The managers always adopt the best strategies for their products, so there are some parts of the strategy and parts of the model that are never going to be chosen thereby the decision is based on the optimal one. For example: choosing futbol team to support: all are almost the same among the same country and league, because all had the same probability to win a tournement. Age and other characteristic of an individual are also random assigned, being a perfect example of random variable.
Deciding between models that are at once almost the same also is rv: pricing at the beginning or with in-apps. However, the definition of this alternatives has to show that too. Other methods are there to reduce this correlation of individuals.
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Bradlow 2017 JR***: Cool paper of what big data is, how it is growing, and what was and what is going to be the impact for the retailing industry. Main ideas here: IOT and apps, are tools to produce data, and high quality data: such as location data. Combining this new data to information of customers of retailing allowing knowing more and more about them, having nice implciation in marketing and sales of the company.
=====================================================================
Chung 2016 JAMS***: adaptive personalization using social networks, not so sure about this one.. I guess it counts specific cases where data is used to create personalization. It explains how this personalization takes place. from data collected to ways to give personal service (recommendation, or just broswer, feeds) some mobile apps include this services. Not easy to identify as user that this service is the justification of a complex data collection.
=====================================================================
Fang 2015 ISR**: using location-based mobile promotions or aslo known as target advertising increases sales of firms, very effective method of using data to do advertising. empirical experiment with control population.
=====================================================================
Ghose and Han 2014 MS***: estimating demand for mobile applications, factors that increase demand.
=====================================================================
Gill 2017 JM***: Some mobile apps are used by organizations and firms to increase the reputation of the brand. Apps used by firm do not generate new sales directly, but increase the reputation of the brand which on second stage increases sales Some types of uses of apps by organizations are mention here: all this apps do not use mechainsms to capture value
======================================================================
Hoehle and Venkatesh 2015 MISQ ***: mobile application usability. what are the factors that increase usability of an app. Some features of the app, and some specific items of the app.. also realted to apps with higher demand. How app developers can increase usability. Interesting for me, the definition of usability in terms of user can perform a task with the app. also it has definition of an app.
======================================================================
Hyrynsalmi 2014 JSS***: sources of value in application ecosystem. Holistic perspective. Definition of app ecosystem and members.
======================================================================
Known 2016 ISR**: addiction to apps, specially to social networks. This is often in some type of consumers. addict users consider less the implications of interacting on the apps when they just want to increase the utility of consuming the app. (sharing data about himself and community with less privacy caution)
======================================================================
Qiu 2017 ISR**: study of independent software developers in the iOS app market. Interesting explanaition of two logics behind a developer: on the one hand, create new apps (perform creativity tasks for fun and to see final output), on the other hand as market imposes aspects of competition, wants to obtain higher market share and profit! store also interact with this type of entrepreneurs.  large firms of developers compete with them, having many advantages over them, which are one man firms in majority.
======================================================================
Tiwana 2010 ISR***: This is a first paper on apps ecosystems. The most important contribution is the identification of this ecosystems as platforms. the authors provide a definition of this platforms.
======================================================================
Urban 2015 MIT SMR*: apps are used to make revenue for firms. best method they found is by providing an app without mech to capture value, a benevolent app that creates value with practical functionabilities, related to specific brand, allows build trust and thereby is an effective instrument to improve e-reputation of firm and increase sales afterwards.
======================================================================
Haan et al. 2018 JM***: switching of devices may have impact on conversion rate. Just using mobile device may have less conversion, but starting the search on mobile devices and then passing to less mobile device (laptop) the conversion is higher, than staying on mobile device, especially  when the risk is high. Nice conceptual model and moderation effect. paper is related to marketing.
https://journals-sagepub-com.dianus.libr.tue.nl/doi/pdf/10.1509/jm.17.0113
=======================================================================
Carare 2012 IER* the rank of sales of today affects the sales of tomorrow in products of electronic market. empricial evidence from apps in app store. nice methodology to deal with ambiguity of rank information (i.e., they have the rank but not the exactly number of units / sales)
=======================================================================
Good example for SBM
Han Park Oh 2015 MISQ ***: the downloads of apps is studied with mobile app analytics to identify how to have higher amount of tries. Now, user data such as measurement of usage, habits and engagement is analized to reach higher level of in-app sales.
Some of the variables used are utility and satiation. modelizing this two through logit model for decision model and regression model for the number of choices
user behavior on the mobile platforms and apps
Data and methods nicely done (econometric model) results
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Note about categories on google play:
Google Play is a leading app market based on Android operating system. Notably when publishing a new app or a new version of the existing app in Google Play, app developers self-select one or more appropriate app categories, however, they are not required to go through the verification process. Hence there are some cases in which app categories reported by app developers are incorrect or inconsistent. To address this issue, Nielsen KoreanClick performed a thorough, manual re-classification task to ensure that a certain app is classified to the single, primary app category. We use the Nielsen KoreanClick's categorization in our empirical analysis. If the categorization of apps were arbitrary, it would be difficult to find meaningful results from the empirical analysis. However, our empirical analysis based on the Nielsen KoreanClick's categorization show highly significant and meaningful results (we will discuss the result in detail in Results section), adding empirical validity to the categorization. Of course, we can adopt alternative categorizations in the proposed empirical framework. In Results section, we also perform an empirical analysis using individual app level data.
 Nielsen KoreanClick classified the mobile content into 14 categories; communication, game, map and navigation, entertainment, lifestyle, personal finance, music and radio, photo, portal, schedule and memo, social networking, utility, video, and combined mobile web activities
=======================================================================
Ghose and Han 2014 MS***:
Demand based view paper!
Structural econometric model to quantify the vibrant platform competition between mobile (smartphone and tablet) apps on the Apple iOS and Google Android platforms and estimate consumer preferences toward different mobile app characteristics.
We find that app demand increases with the in-app purchase option wherein a user can complete transactions within the app. On the contrary, app demand decreases with the in-app advertisement option where consumers are shown ads while they are engaging with the app.
Literature on: user behaviour under product characteristics, only with digital products. (e.g., minutes of phone under pricing settings)
Explanation of apps economics.

We show that demand increases with the app descrip-
tion length, number of screenshots, in-app purchase option, app age, version age, number of apps by the same developer, number of previous versions, cross- chart listing, cross-platform listing, and volume and valence of user reviews. On the contrary, app demand decreases with file size and in-app advertisement option. Older and male consumers tend to be less sensitive to the price of apps than younger and female consumers, respectively. On the supply side we show that app file size is a major cost driver in app develop- ment, but that there are significant returns to scale in app development. Cost decreases with in-app purchase, in-app advertisement, app age, and age restrictions. Compared with lifestyle apps, games, social, and utility apps have higher marginal costs, whereas media apps have lower marginal costs.
=======================================================================
area medicine -
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4029126/
Mobile Devices and Apps for Health Care Professionals: Uses and Benefits
C. Lee Ventola pharmacy and therapeutics 2014
=======================================================================
Rebecca Balebako ; Lorrie Cranor
IEEE Security & Privacy 2014
Improving App Privacy: Nudging App Developers to Protect User Privacy
Smartphone app developers make many privacy-related decisions on what data to collect about users and how that data is used. Based on interviews and a survey of app developers, the authors identify several hurdles preventing app developers from improved privacy behaviors. These include the difficulties of reading and writing privacy policies as well as privacy not being their primary task. 
app developers can use third-party developers to perform analytics and ads, to add extra services to their apps. This third-party tools can retrieve user data through the app, without the entire knowledge of the app developer. The developers usually don't read the privacy policy of this tools because they are quite difficult to understand, and in many cases are only to obfuscate the understanding of the content. app developers, in turn, don't give too much attention in writing clear pp for their apps, while some just copy and change wording of someone else's pp, or others simply just don't have any pp. Privacy is not a priority for app developers.
The author interviewed developers to understand their behavior.
=======================================================================
Carlos Jensen, Colin Potts
Privacy policies as decision-making tools.
what is the purpose of pp
problems of redeability of pp due to the complexity of the writing of it
problems with access to pp
problem with notification in case there are changes.
=======================================================================
What catalyses mobile apps usage intention: An empirical analysis
JJ Hew
What are the types of mobile apps, in the case of apps of Google Play we say they are native apps, because they can only be used on specific operating system. e.g., iOS. There are also web apps, and hybrid apps. the web ones, have the structure of an app but they are launched on the web and people can have access through an url, without nithing any type of app market.
Mobile apps are defined as software or a set of program that could be executed to perform certain tasks for users on m-devices
The factors that determine the uses of a certain app. the author tested variables such as price, performance expectancy, effort expectancy, facilitating conditions, habit, social influence, and hedonic motivation.  to see the results refer to the paper, in general, he did cross-sectional to find the factors determining behavioral intention to use a mobile app.
=======================================================================
The Privacy and Security Behaviors of Smartphone App Developers

Balebako et al. 2014 Carnegie Mellon University 

smartphone app ecosystem: the popularity of app store + relationship between add revenue model and data collection

user's perception to smartphone privacy: users want to be more protected, they don't understand the reason for permissions, and they don't want to allow apps to take that data. a privacy policy is not easy to read and sometimes they are not precise in the content. 

public policy and information to app developers to respect the privacy of users there are 

interview to 13 dev and survey to 228 with questions about data collection, permissions, privacy policy, third-party tools

=======================================================================

Mobile application market: A developer's perspective
Adrian Holzer and Jan Ondrus
what are the main actors in the development of mobile apps. 5 different operating systems.
the process of mobile app distribution. First, the developer uses development tools to build its mobile application. Second, the developer publishes its application on a portal, from which the consumer can download the application onto its mobile device.
this process explain the participation of third-party member that intermediates developers and consumers. this is also a two-sided market that experiments of network effects.
app developers can use SDK software development kids
classification of these operation systems
Google is a two-sided market or a platform (depending on the point of analysis) Open technology: platforms grant developers access to all or parts of the SDK and OS source code, there is no central architect responsible for the platform. It has a centralized portal, Google Play is the main portal on which all applications are published. this portal is a single point of sale. This operating system can be used in diverse devices.
Being on centralized portals implies big base of consumers, low entry costs, limitations on content.
Being on open technologies implies low development costs of applications. Device variety increases freedom for developers.
=======================================================================
Sources of value in application ecosystems
Hyrynsalmi Seppanen Suomimen 

Mobile application ecosystem is an interconnected system comprising an ecosystem orchestrator, mobile application developers, and mobile device owners, all of whom are connected through a marketplace platform.
using the concept of amit on sources of value creation in e-businesses: efficiency, lock-in, complementary, and novelty, the authors recognize on the mobile app ecosystems what value is created.
=======================================================================
The Quantitative Discovery: What is it and How to Get it Published
Bamberger, Peter
Ang, Soon
Academy of Management Discoveries 2016

the application of quantitative approaches to describe and examine organizational problems, anomalies, and management-related phenomena lying beneath the radar may serve as a critical means of laying the groundwork for theory generation.

providing empirically driven insights to guide theoretical development along the way

quantitative data collection and analyses in AMD articles are undertaken for the purpose of revealing, describing, and diagnosing interesting phenomena that are poorly understood, as distinct from purposes in other journals of testing hypotheses or filling gaps in established research areas.

Indeed, as noted by Van de Ven and his editorial team in an earlier FTE (2015), discovery is grounded on the logic of abductive reasoning elicited by the observation of “astonishing phenomena” or empirical anomalies (Hanson 1960: 104). Capturing these phenomena or anomalies in the form of quantitative data, the process of discovery is therefore structured around activities aimed at inferring preliminary theory from numbers and numerical patterns, and using quantitative findings to modify and enhance the predictive utility and explanatory potential of such theory (Kulkarni & Simon, 1988).

 investigators identify an interesting anomaly and then use empirical observation to “tweak” it and learn more about its properties and effects.

In this age of access to big datasets often consisting of thousands of observations, statistically significant findings are easy to obtain on minute magnitudes of effects that are not practically important

quantitative data are likely to be the “data of choice” for a number of discovery-oriented research activities and objectives. One of these activities has to do with classification. Quantitatively driven taxonomies and classification systems provide a basis for description on the basis of phenomenological distinctiveness. Quantitative data can facilitate the identification of repeating patterns and commonalities, as well as the preliminary testing of hunches about the clustering of phenomenon and the distinctions among emergent types.

Applying such tools to “big data” may allow scholars to detect patterns of emergence that are only “visible” in large numbers and impossible to detect on the basis of even the most sensitive qualitative techniques

Following the suggestions from … on how to make discoveries off of quantitative data,

follow up on hunches or ideas about discovered anomalies to test these new insights

 examine data patterns with an eye toward potentially important anomalies

Researchers should consider drawing on big data to make quantitative discoveries, particularly when the domain of discovery involves phenomenon with a low base rate or low sensitivity

we suggest that scholars take advantage of methodological innovations to uncover phenomena previously deemed inaccessible or difficult to measure or capture

scholars may use unconventional quantitative methods to statistically uncover critical latent patterns

========================================================================
Construct Validity Research in AMD
Bamberger, Peter
Academy of Management Discoveries 2017

Academy of Management Discoveries (AMD) was established by the Academy of Management with a mandate to surface phenomena that are poorly captured and explained by existing concepts and theories

Certainly no phenomenon can be fully understood until it can be measured, and an understanding of the relationship between it and other constructs is contingent on an understanding of how such latent or unobservable factors are indicated by observable and measurable factors (Edwards, 2003; Schwab, 1980)

shifts in the nature of work, human relationships, technology, organizations, and organizational environments, there seems to be a never ending flow of new phenomena that we as scholars need to account for and explain and thus need to be able to characterize and measure

The fact that a phenomenon is new and poorly understood is already interesting in that everything we can empirically uncover about it is a “discovery”

no matter how interesting a phenomenon may be, until it can be accurately and reliably measured, our ability as scholars to understand such phenomena, explain their origins and demonstrate their implications for management is extremely limited

Although our field has no shortage of overlapping constructs, as new phenomena in management and organizations are constantly emerging, it is our obligation as management scholars to account for them and incorporate them into our knowledgebase
========================================================================
AMD —Advancing Discoveries Through Empirical Exploration
Van de Ven, Andrew H.
Academy of Management Discoveries 2017

AMD welcomes exploratory studies at the pre-theory stage of knowledge development, where it is premature to specify hypotheses but where plausible hunches are needed to guide future theory and research.

articles published in AMD use an abductive process of reasoning to diagnose poorly understood phenomena and conceive of plausible conjectures about them

In addition, each article contains hyperlinks to a variety of multimedia content such as interview excerpts, video clips, pictures, illustrations, and dynamic simulations. The goal of these and other media innovations is to not only use digital media to illustrate and engage in scholarly conversations but also advance social scientific knowledge in ways that transcend the limits of paper text. It is not about making pretty pictures—it is about doing better social science.
========================================================================
26
Finding New Kinds of Needles in Haystacks: Experimentation in the Course of Abduction
Mueller, Jennifer
Academy of Management Discoveries 2018

Experiments in management needs a priori, specific focus to start. Abductive research is more exploratory.

Abductive: a “process of reasoning from data to an initial hypothesis” (Behfar & Okhuysen, 2018: 1).

often beginning with an intriguing question that cannot be easily answered on the basis of extant theory or research, followed by demonstrating a series of plausible relationships in the data and ending with an effort to provide a broader framework to potentially explain these tentative relationships

experimental projects are defined as studies conducted in the laboratory, field, or online, which randomly assign participants to a given treatment condition

In my view, when it comes to engaging in abductive inquiry, experimental projects are especially helpful to employ when needing to (1) surface emergent or poorly understood phenomena using a more precise operationalization than used in prior research, or (2) surface relationships that are surprising because the authors can more precisely rule out known and expected alternative theoretical frameworks

Given this, one might ask how to go about designing and writing up an experimental study for AMD? First, instead of starting the article by proposing a set of a priori hypotheses, researchers might pose a research question fueled by the identification of discrepant findings or a conceptual puzzle with significant theoretical and/or practical importance. In sum, the authors posed a research question that current theory could not adequately answer and then offered a compelling justification for their inquiry, suggesting that relative to extant conceptualizations of status, relational comparisons might better reflect people’s psychological experience of status.

Second, after proposing an important research question, AMD experimental articles typically offer multiple experiments or studies, with each building upon the one before it while still tackling a different aspect of the central inquiry.

The final step of writing an article for AMD is perhaps the most impactful for future work: write a general discussion section summarizing and integrating the observations from all of the experiments reported.
========================================================================
27
AMD—Clarifying What We Are about and Where We Are Going
Bamberger, Peter A.
Academy of Management Discoveries 2018

AMD: empirical exploration in management

Data-driven approach

data-driven approach taken to surfacing phenomena and/or providing robust and parsimonious “first suggestions” for them—plausible insights into the nature, antecedents, and consequences of such phenomena, as well as the new or transformed theoretical frameworks required to make sense of them

abductive reasoning  “is an inseparable, indispensable, and valuable approach linking the development of explanation and the testing of resulting hypotheses to advance theory”  , Okhuysen and Behfar (2017)

theory Provides assumptions to be challenged and frames anomalies to be explored and suggests the variables on which to sample

abductive reasoning is the weakest form of reasoning of the three, allowing the researcher to emerge with only a plausible conjecture and some insights into what this conjecture might mean for the development of new or alternative conceptual frameworks (Shapira, 2011) and down-the-road theorizing. Although abduction offers a logic for considering conjectures about complex phenomena, it does not produce simple or clear answers

Locke, Golden-Biddle, and Feldman (2008: 907) note that “deduction proves that something must be, induction shows that something actually is operative; abduction merely suggests that something may be.”

scientists cannot confirm hypotheses deductively when knowledge is limited and fragmented, because experiments will likely fail and the results provide no indication of where else to explore.” It is in such situations that we enter the realm of empirical exploration, digging deep into patterns embedded in our data to generate the tentative and fallible conjectures that may eventually lay the groundwork for innovative theorizing and subsequent hypothesis testing

Abduction may be applied in a wide range of circumstances in which we encounter a phenomenon or patterns of relations that challenge extant knowledge.

Types of abduction:

The first type, which I refer to as exploitative abduction, involves the systematic collection of “facts,” followed by an attempt to identify a framework which explains the pattern of facts identified researchers applying 2018 Bamberger 3 exploitative abduction engage in a process of elimination, eliminating concepts and theories that fail to connect the (observed) dots. Accordingly, when engaging in exploitative abduction, we first collect as much information as we can about the phenomenon of interest. We then contrast what we observe with what we would expect to observe was some general framework or theory to apply, ruling out explanations that fail to account for the configuration of evidence we observe. Finally, by demonstrating that some general rule or theory which could not be dismissed explains a lot or most of what we observe, we conclude that this general rule may be what Lipton (1991: 61) calls the best available or “loveliest” explanation. There are some explanaition that theory suggest, but we are not sure!!

The second type of abduction, which I call exploratory abduction occurs when the researcher is confronted with puzzling facts, but unable to cleanly apply a theory or theoretical perspective to readily explain them, uses the pattern of results to conceive a plausible explanation, or at least identify the criteria that an explanation would have to meet to be plausible. Engaging in exploratory abduction, the researcher must herself conceive the general rule and use the pattern of findings to argue for its plausibility. As with exploitative abduction, here too the steps taken to move toward a plausible conjecture involve contrastive reasoning, comparing what we observe in fact to what we would expect to find were some extant theory to apply Such contrasts allow the researcher to narrow the range of plausible explanations, providing a grounded basis for the development of conceptual frameworks (Shapira, 2011) and downthe-road theorizing. In quantitative abduction, these contrasts are executed on the basis of sensitivity or robustness analyses where the investigator openly plays with the sample or model specification (i.e., including or excluding controls and/or particular sets of observations) in an effort to “rule out the usual suspects,” or on the basis of experimental manipulations designed to do the same.

, both types of abductive inquiries may be initiated on the basis of a common set of triggers. Based on my experience in handling manuscripts for AMD over the past 3 years, I can identify two main types of triggers for abductive reasoning. The first is deliberate, whereas the other is more opportunistic. Deliberate abduction is driven by an interest in understanding a phenomenon that although commonly observed, cannot be readily explained by extant theory. By contrast, opportunistic abduction is evidenced in those situations in which, in the process of conducting an inquiry into some intended research question (e.g., how light affects productivity), we stumble across findings that are surprising, counterintuitive, and/or anomalous to extend understandings (e.g., productivity increases as we approach complete darkness). an observed, robust pattern of relations is not easily explained by an extant theory or theoretical perspective

In this context, inquiries grounded on abductive reasoning are important to our discipline for four main reasons. First, such inquiries offer a critical means by which to surface anomalous relations, stylized facts (Helfat, 2007), and empirical regularities (such as a link between smoking and cancer). This is important according to Hambrick (2007: 1348) because by doing so, “subsequent researchers can then direct their efforts at understanding why and how those facts came to be.” In other words, abduction precedes research in the hypothetico-deductive tradition by identifying and describing the phenomena worthy of study and laying out the parameters of plausible explanation that can then be integrated into some theoretical framework subject to testing and confirmation.

inquiries grounded on abductive reasoning are important in that they are grounded on pragmatism (rather than positivism) and are open and transparent, therefore, offering an important response to questionable research practices. In abduction, there can be no HARKing because there are no a priori hypotheses. Indeed, although one can argue that abduction is all about HARKing (i.e., the data drive the inferences), it is critical to remember that the objective is to infer plausible yet fallible conjectures from empirical realities, not to confirm them.

Surface significant patterns of occurrences, using any number of empirical approached including rich description, quantitative construct specification, and or empirical taxonomic analyses.

Identify and explore surprising relationships using rigorous qualitiative and or quantitative.

Offer empirically driven insights into and/or a plausible resolution of critical anomalies and discrepant findings. Situation presents a need to explore and discover a new and plausible explanation,

Authors of such papers should ensure that their manuscript (a) offers scientifically rigorous evidence of the relationship’s consistent (e.g., evidence of the robustness of the relationships across methods and/or samples) and non-spurious nature (e.g., reasonable assessments of robustness and sensitivity), (b) provides preliminary evidence about or at least speculates on the mechanisms underlying the phenomenon or relationship, (c) explains the significance of these findings to management and organizational research, and (d) lays out a strategy for further exploration and/or downstream theorizing