Week4(2/6)
The Data Walkshop and Radical Bottom-up Data Knowledge(Allison Powell)
Top-down approach
Researchers have certain kind of assumption that...what kind of data they are looking for?
Researchers have certain kind of assumption that...what kind of data can be retrieved from research?
Bottom-up approach(Data walk)
“Construct a narrative for how they define and critique data in place” through:
walking/observing/reflecting/remarking
“Bottom-up approach allows us to attend to who is asking the questions about big data, and, further, lets us think about how data gets to be ‘big’ in the first place, who asks the questions that make it big (in size as well as importance) and how one might ask different kinds of questions.”
Subjectivities from interdisciplinary encounters
"Participants, with their specific roles to perform, observed the world but also experimented with playing at observing the world. Rather than assuming that ethnographic observation collects truthful observations, this perspective points out how much our situated knowledge is constructed through our experience, and invites us to shift that experience, and to be reflective about what the shift produces."
Questions
1.We are taught to be objective and make no assumption when we work with data. But also being completely objective is impossible as long as we take a certain type of role in the community. Powell suggests the interdisciplinary team would help us balancing those two perspectives, but what perspective can we have, if we're working alone or with limited resources?2. Even though we gathered insightful data from data walk, we do not always have access to public data and chances may be limited to relate collected data to a decision making in a broader context. How can we involve decision makers in a bottom-up approach?
----------------------------------------------------------------------------------------------------------------
Working Ethnographically with Sensor Data(Dawn Nafus)
1. Describe a straightforward way that this research practice involves for those who might be curious about adapting it.“’ordinary’ people do in fact have something to say about which patterns matter and which pattern do not. Big data is not just about esoteric calculations that require machine learning algorithms.”
2. Contribute to the discussion about the entanglements between materials, sociality, and method.
Research method example 1:Ethnomining
Research question:
Market intelligence about how laptops are usedChallenges:
-Find less meaning in analyze from a demographic viewpoint.-Incorporated Ethnomining and involved participants into research.
Keys:
“balance between giving participants enough of an explanation of how the graphs worked so that they stood a chance of interpreting them at all, and giving not so much that we ended up with flat responses, or telling them what to think.”Research method example 2:Caregiving
Research question:
What specific aspects of stress were causing the most trouble?Challenges:
-Several variables (Sampling rate, Battery and data storage, Burden on participants etc…)Keys:
“The ‘full story’ is not what made it interesting; the fact that it came in layers, with different parts remembered in different ways, is what gave it meaning.“These nuances became visible only by making room for, and valuing, human ‘error’”
Questions
1. I understand intervention such as interview is the key to insightful analysis. However, I feel this method has a risk that researchers can potentially mislead hypothesis by intervening arbitrarily.This may unlikely happen in marketing research in the given example, but what if this research was census run by a government which wants to lead a certain type of result? How can we guarantee the research result reflects community peoples’ voice?
2. It would be almost impossible to get rid of every variables sensor data, so we need to some criterion to secure the validity of data. I usually argue the validity of numerical data helped by statisticians with their viewpoints. How a researcher like me without that background verify that certain data (and arguments based on the data collected) is valid? Is there any criterion?
----------------------------------------------------------------------------------------------------------------
Nietzsche, Genealogy, History(Michel Foucault)
Genealogy of genealogy
“The analysis of descent permits the dissociation of the self, its recognition and displacement as an empty synthesis, in liberating a profusion of lost events.”“it is to discover that truth or being does not lie at the root of what we know and what we are, but the exteriory of accidents”
Objective: Entstehung/Herkunft: origin. Herkunft is equivalent to stock or descent.
-The ancient affiliation to a group, sustained by the bonds of blood, tradition, or social class.-
Effective history VS Traditional history
→Reframe history from a seamless continuity from the past, to a profusion of entangled events.Effective history
-Deals with events in terms of their unique characteristic, their most acute manifestations.Traditional history
-Dissolving the event into an ideal continuity“ It (“Effective”) history will uproot its traditional foundations and relentlessly disrupt its pretended continuity. This is because knowledge is not made for understanding, it is made for cutting.”
“But the true historical sense confirms our existence among countless lost events, without a landmark or a point of reference.”
1. I felt approach and perspective of genealogy something in common with ethnography. They both focus on the dynamics of complex context and micro perspective. Is there any academic interaction?
2. Foucault argued successes of history belong to those who are capable of seizing the rules. He also described what constitutes success of history, but is there any concrete example? Does this equal to Winston Churchill’s saying "history is written by the victors"?
