Theme 6: Reflection

I see case studies a necessity rather than simply a powerful tool in certain areas of media technology research (for example studies of communication, social media etc.). The objects of interest for study within such areas can be hard, if not impossible to recreate in a controlled environment, making real world examples a necessity in order to have something to examine.

On a more personal note, the close real-world connection of case study makes this method more interesting for me. I only have experience of case solving, as I haven’t had the chance to do proper case studies yet. But even though these two have different goals I see similarities. Both are based on a real world problem or event and use existing research to look at the subject, but in one case, the goal is to propose a solution while in the other, it is to find answers to questions such as “why?”, “how?” and so on.

The texts I’ve read for this theme gave me a concrete example of another difference between the activities mentioned above, namely the tied in quantitative research. The case solving I’ve been part of were all given a very limited time span (and as far as I know, this is applied to most case solving’s), meaning that was no time to confirm one’s hypotheses with collected quantitative data.

It’s hard to tell now, but I have got a feeling that this theme will have been of most use for me in my work with the master thesis essay. The first paper I wrote about in the precious post (regarding Facebook and personal security) raised a question mark in regard of the target group sampling in said article. It will remind me to be extra careful when I formulate the sampling group for the quantitative data collection.

Theme 6: Qualitative and Case Study Research

Part 1

I chose to read ”Facebook and online privacy: attitudes, behavior and unintended consequences.”, published 2009 in the 15th issue Journal of Computer-Mediated Communication. The article uses literature research to present three frameworks which are in turn used to present four hypotheses along with four research questions.

Further on, data collection from university students along with eight interviews were conducted in order to confirm or falsify the hypotheses.

Broadening the age span as well as social groups and maybe even countries for the questionnaire would’ve yielded clearer results in my opinion. The authors refer simply to users and young adults in general, while the sample group was taken from a single college. Different ages and social groups would’ve allowed for comparison of how different life aspects affected peoples attitude towards online privacy and in turn, the conclusion of the questionnaire data. 

The benefits of literature research is the analysis and depth it often provides within a research field. By investigating the number of available literature within the given field in advance may also indicate whether there is enough material for drawing your own conclusions (or it may indicate that the topic in question already has been covered from the intended viewpoint). Properly chosen literature also allows a writer to focus at drawing new conclusions based on previous results, instead of having to compile them from scratch. Fortunately, the number of literature written is more than enough to cover many researching needs of writers.

It’s also a practical research method, as no coordination of interviewees, forms or other ways data collection is needed.

A limitation of said type of research is that it can get out of date when newer findings are made, as new literature refer back in time (i.e. old literature) whereas the other way around is yet to be seen. However, it’s a limitation which is dependent on the researcher, as it’s up to the researcher to make sure that he/she has the most up-to date sources. Writing a paper on an internet-related topic using literature from the 90-s would most probably result in many faulty conclusions (it may be a far fetched example, but it illustrates my point).

The benefits of interviews is that they offer an in-depth view on a subject from the respondents perspective. It allows to get answers to the ”why” and ”how”-questions which are necessary to get a better understanding of a topic. Additionally, an interview situation allows the interviewer to ask follow up questions to certain answers, unlike quantitative methods, which opt out the flexibility and in-depth possibilities for the benefit of a greater amount of data.

As for practicality, the interviews have the downside of having the need to be planned before being executed, the difficulty of which is determined by the schedule of the subject to be interviewed (which in turn isn’t entirely in the hands of the researcher).

Part 2

A case study is, in short, an examination of an event using a set of different analytical tools, often to get an answer to questions or to formulate a theory applicable to similar cases.

The article I chose is ”The tweets that killed a university: A case study investigating the use of traditional and social media in the closure of a state university”, published in the journal of Computers in Human Behavior issue 29, 2013. As the title suggests, the authors seek to investigate a single event and not a more general phenomenon, which explains the number of cases examined in this article being limited to one. As the question under investigation is quite specific and the event is one in it’s kind as far as I know, I don’t see the need for additional cases to be analyzed.

The article starts off by a brief explanation of the event, as well as a number of sub-topics, which are referred to later on in the data collection and analysis for validation. The information for the sub-topics was gathered from external research and not conducted by the authors themselves. As for the data-collection, the authors used two main methods: questionnaire and data mining analysis. Besides questions for the main topic, the questionnaire also contained questions used to confirm or disprove previous findings described in the sub-topics (in the end confirming them). In the end, the data mining analysis was coupled with the results from the questionnaire to further strengthen the findings. As the purpose of the study wasn’t as much to form a theory as to study a single event, the result wasn’t a model of any kind, but rather a discussion of results and proposals for similar situations in the future.

Theme 5: Reflection

One, in my opinion, interesting aspect of research in the media technology field is that of a new technology’s application of usability by the public on a broad scale isn’t entirely determined by the producers/researchers, making it hard to predict the “correct” direction to go during the design process (unlike, for example in the field of medical research, where people tend to use products according to directions). Obviously this doesn’t necessarily apply to all cases, but it’s nevertheless a possibility.

Twitter is one example of what I mean. It was initially developed and launched primarily as a tool to let friends and family up to date of your thoughts and what is going on. Even early on there were internal disagreements whether the focus should be on the tweeting individual (i.e. “Drinking Coffee at Starbucks” and such) or on his/her surroundings (i.e. “There’s a fire at Starbucks on Central Station” and such). The Arab Spring, where Twitter was used as a channel for the protesters to make themselves heard, showed that it was much greater that both of the previously mentioned focuses. The hashtag was dismissed early on as too “geeky” for the average Twitter user, but can today be seen everywhere in social media. Companies have taken to this technology as a tool to strengthen their relationships with their customers and make the brand feel more alive (one of my favourite examples of this is Oreo’s tweet during Superbowl 2013, who during the infamous blackout tweeted a picture of an Oreo accompanied by a text saying “You can still dunk in the dark”).

Perhaps this uncertainty may be solved in the future with new design methods. The development in the design process started with participatory design, which integrated end-users into the prototyping phase at the end of the whole project. This later led to a user-centered design, where the user experience was in focus and the end-user’s participation was moved to earlier stages in the design process. Further down the road, service design emerged, which focused on how the product of the research would fit into the end-user’s life rather than on the isolated interaction with the said product. Finally, we have the human-centered design, which makes the switch to a societal view instead of an individual one when designing the product in question.

Or perhaps it’s something that can’t be solved thanks to the unpredictability that is human nature.

Theme 5: Design Research

Both concepts described in the articles were interesting, and one key difference I noticed (besides the obvious difference in the prototypes) was how one type of prototypes could be tested more or less in the way it’s going to be used (as far as I could tell at least).

The three robotic prototypes described in ”Comics, Robots, Fashion and Programming: outlining the concept of actDresses” were/are most likely going to be tested in similar conditions as when they are going to be used by consumers. In the case of the tactical information system, the main problem it was trying to solve was how to present information  on the go without distracting the user from his/her surroundings (which is the case with visual information), which i interpret more or less as being able to take in information while doing something else. During the tests described in the article though, users were focusing at the information at hand.

I think this difference exemplifies the limitations of prototypes depending of the field of research and stage of development of a project as well as the type of role they may play. In fields where several well established technologies are combined in new ways, prototypes may show a clearer picture of the end-results. In less explored fields, prototypes may on the other hand simply point if the research is headed towards the right direction or not. If we take the soccer example: a prototype field-test with the same prototype may have given different results than the one described in the article, but it’s on the other hand quite understanding that recording the user experience in the field would’ve been rather complicated compared to controlled tests in a room/laboratory (or whatever facility one may choose). This is one example to why it could be necessary to develop a proof of concept prototype.

Theme 4: Reflection

I recently finished reading ”Freakonomics”, a book which is based more or less entirely around quantitative research. The authors, Steven D Levitt & Stephen J. Dubner, use data to answer questions such as ”How is The Ku Klux Klan like a group of real estate agents?”, ”Why do drug dealers still live with their moms” and ”What do school teachers and sumo wrestlers have in common?”. It was interesting to see how quantitative methods could bring answers to such rather odd questions and the power of explanation that statistics have.

It can, on one hand, be argued that quantitative research is more objective due to it’s close connection to the real world (after all, quantitative data could be viewed as the real world from a certain perspective transformed into numbers). On the other hand, due to the need to have the correct complementary data, the lack of said data may lead to loss of objectiveness. For example, simply saying that there are less plane-crashes than car-accidents will give the illusion that traveling by air is safer than by road. Taking into account the amount of air and road traffic would instead show that neither of them is a lot more dangerous or safer than the other.

While this example is oversimplified, and both cases are technically objective (after all, numbers don’t lie), the question of the level of objectiveness of quantitative research turns into a philosophical one. If I’d like to convince my audience into believing something, I can choose the right complementary data to make my case. I do realize that this isn’t a scientific approach, nevertheless it’s still a question of objectivity.

Theme 4: Quantitative Research

My paper

The paper I’ve chosen for this topic is ”The Long-Term Effects of E-Advertising: The Influence of Internet Pop-ups Viewed at a Low Level of Attention in Implicit Memory”, published in Journal of Computer-Mediated Communication 7th October 2013.

Like the title suggests, the authors look at how pop-ups affect our attitude towards product brands. The starting point of the methods used in this paper is a set of theories covering the implicit-effects that different types of brand exposures have on the target audience over time. The different exposure-types were:

1: The brand logo by itself.

2: The brand logo with the product represented in words.

3: The brand logo with the product represented in images.

Quantitative methods were then used through three phases in order to determine which of these three exposure-types were the most effective in achieving the following effects on the test subjects: increasing the experienced quality of the brand, the likelihood of purchase, increasing the response time of a negative answer when associating the exposed product with a cheap brand. The test subjects were 398 university students between the ages 19 and 21, amongst which the gender ratio was 49/51% male/female. Students were chosen because the research group wanted the test group to be easily contactable.

The first phase of the quantitative research was conducted by having the subjects visiting a fictional health-related website of professional quality, where they were exposed to a pop-up of a made up brand. Since the goal of the survey was to measure the implicit effects of popups, the subjects were told that they were participating in a poll for the French Ministry of Health.

In the second and third phase, subjects were called in again, but this time with different persons holding the surveys in order to eliminate any connections with the first phase. These surveys were completely focused on brand research and contained the brand featured in the popup as well as three filler brands. The subjects answers as well as response times were recorded. The second and third phases were conducted a week and three months respectively after the first phase.

After the surveys were completed the data was analyzed. The conclusion in short was that popups containing images affected the attitude of persons towards the better

After mainly having been exposed to quantitative methods in the form of polls and such, I learned how quantitative research can be planned and followed through in more advanced ways than straight forward questions. 

The main methodological problem I saw in this research was the missing accountability for variables such as different types of images, products, users and so on, but the authors do note this in the concluding chapter of the paper. One way of improving the research while keeping the way it was conducted (i.e. focusing on a single brand) would be add questions which aim to explain certain feelings towards brands which would include the ”why” and not only the ”what”, which on the other hand could risk to reduce the number of participants due to the increased time needed from each participating person.

The Physical paper

The medical text, like the article chosen by me for this assignment, aims to support existing theories rather than establish new ones. It shows that qualitative and quantitative research aren’t better or worse than one or another, but each has it’s purposes. While qualitative research may aim to propose new theories, quantitative research takes over where the first type ends, proving or disproving said theories through data. While qualitative research says why physical activity may or may not improve ones resistance to the infections discussed in the paper, quantitative research puts the thesis to the test by saying whether it’s likely or not on a broader scale. Another point that the medical paper clearly illustrates is how variables have to be taken into account when analyzing the research material (in this case: possible pollen allergies, risk of infection from children or coworkers and other factors not related to the question at hand).

One of the limitations of quantitative research is that the time needed to collect a certain amount of data is less controllable than with qualitative research. While a good choice of methods of collection may speed up the process, there is no guarantee that it will yield the desired results within a given timeframe (with the obvious exception for when time isn’t an issue).

Further on, the sampling sources need to be chosen carefully in order to yield valid results in a given field. They also need to achieve a critical mass in order to give sufficient support to the research in question.