Answers to my comments

Brainy Bad Luck Brian17 december 2013 04:13 
Heya! I noticed that you commented in your post about the samples taken regarding ethnicity and gender. My question to you would be do you think that studies like these with a "skewed" sense of sample can really count on the results produced? Is it enough in their paper to point this out but still go ahead with the study? SvaraRadera

Svar Olle Bergendahl18 december 2013 02:31 
I think it depends on if there is reason to believe that the two groups (male and female) might behave differently or answer the questionnaire differently. If this was a study about something like human length (in centimeters/inches) a sample skewed towards females would of course affect the outcome of the study. But in this case, is there really reason to believe that females behave differently? From what I understand you do not really "play" Second Life as yourself, instead you have an imaginative avatar. A female player might decide to play as a male in-game, and a male player might decide to play as a female in game. So maybe the gender does not matter that much... On the other hand, I kind of agree what you suggest in your second question. It should not be "enough" to just point out flaws in the method, and then just go ahead and present the results as reliable. Of course it is a good thing that the authors are aware of any possible weaknesses/mistakes in their method, but the best thing would be to not make those mistakes in the first place.
-------------------------

Ekaterina Sakharova26 november 2013 06:51 
I do not agree with Sutton who says that references, data, variables, diagrams and hypotheses are not the theory. I think that references, data etc. are not the theory itself, but I am sure that data and diagrams are important parts or ingredients of theory. And I always though that hypotheses are the base of theory. SvaraRadera

Svar Olle Bergendahl4 december 2013 05:27 
I think you are correct when you say that data, references, etc. are an important part of the theory. Without underlying data there is nothing to base the theory on, and such a theory is more or less useless. A hypothesis might be the base of a theory, but in my opinion a hypothesis is not based on raw data but on "belief". A hypothesis is want you think the outcome of an experiment might be before you conduct the experiment. In other words I guess you could call a hypothesis a guess that is based on your previous experience, but not based on data/facts about the current case. Radera
-----------------------

leah22 november 2013 01:54 
I like how you looked up the authors on wikipedia. It's important to find out what sort of perspective they were writing from to make sure we do not mistake their opinion for the majority. It seems much clearer why they did not appreciate American society due to their ideological views, while others maybe would have embraced it. SvaraRadera

Svar Olle Bergendahl4 december 2013 05:18 
Thanks for your comment. Yes, reading on Wikipedia gives you a completely different picture of the authors' opinions. I'm not sure but maybe they were quite alone in having such a negative view of the culture industry. I think this is something that should have been discussed during the seminar.
--------------------------

Brainy Bad Luck Brian20 november 2013 03:50 
Hej Olle, I was wondering what you would think Adorno and Horkheimer would think of the new media technologies today? Looking at other people's blogs it seems like the general accordance is that they would be horrified and dismayed at the point in which the culture industry has gotten to now. Is there any redeeming features in your opinion of new media today? SvaraRadera

Svar Olle Bergendahl4 december 2013 05:14 
I agree with you to some extent, but I also think that there are benefits of the current media technologies. It is true that the culture industry in general is more about money/capitalism than ever before, but I also think that the Internet have helped to redistribute some of the power concentration. One of Adorno and Horkheimer's main arguments is that the culture industry concentrates power to a few powerful people. To me it is clear that the Internet is doing the opposite, giving "ordinary people" a way to be seen by a large audience and convey their own message or opinion. Examples of such services are blogs, Twitter, Facebook, Youtube, etc. A more concrete example would be the Arabic Spring.
-------------------------

MatteoCampostrini20 november 2013 08:07 
I see your point about the purpose of myth, mainly is a tale to explain phenomena we were not able to explain and still there are many mysteries in our universe. For instance any ancient civilization has its own creation myth, nowadays we know about the Big Bang, scientist perfectly described backwards physical forces up till what they call "the Planck epoch", 10^(−43) -(0, 43 zeros and 1)- seconds after it but still we don't know what happened before. Why do you think we do not use myth anymore to explain this? SvaraRadera

Svar Olle Bergendahl4 december 2013 05:15 
I think some people still like to use a "myth" to explain the creation of the universe or what happened before the universe was created. If we look at the States we see that some people still believe that the Earth was created by God roughly 6 000 years ago. But I guess that you are referring to the scientific community and why they do not try to explain what happened before the Big Bang. I believe that what happened before the Big Bang is "outside" of the reach of science. The basis of science is that it has to be based on observations, experiments and falsifiability. If a scientist makes a claim of something that happened before the Big Bang, it is impossible for another scientist to try to verify or dismiss that claim. Therefore such a claim is not scientific, and there is no point to try to discuss such events in a scientific manner. There is a very good example of this coined by the philosopher Russell (that we read in Theme 1). It is called Russell's teapot and you can read about it here: http://en.wikipedia.org/wiki/Russell's_teapot
----------------------------

VictorO15 november 2013 03:40 
One thing that I came to think of after reading your post is that although "true knowledge" cannot be derived from "mislead" knowledge in Russell's meaning, can it the knowledge itself be rejected since from the perspective of one individual truth can be experienced on way and from another can be experience another? SvaraRadera

Svar Olle Bergendahl4 december 2013 05:15 
That is a very interesting thought! I guess you mean that in some cases there are no real "truth", since each person could interpret reality differently. Any knowledge based on such a truth might therefore have to be rejected if we look at it the way Russell does. Even though it might seem reasonable to reject such (subjective, or whatever you want to call it) knowledge it leads to problems in some cases. As the theory of relativity states, even "hard" facts/measurements such as time, distance and velocity depends on the perspective of the observer. I guess that if we apply Russell's thought to the theory of relativity we cannot obtain any knowledge that is based on facts/data of distance, time or speed. This causes some weird paradoxes, such as http://commons.wikimedia.org/wiki/File:Fotonklocka.svg http://en.wikipedia.org/wiki/Ladder_paradox http://en.wikipedia.org/wiki/Bell%27s_spaceship_paradox

Theme 6: Qualitative and case study research (Post)

In this week's seminar (Monday) we first presented the papers that we read and then discussed pros/cons of qualitative methods. We started off by briefly presenting our own papers, and then choosing one paper to present to the whole group. After some discussion we decided to pick my paper, The psychological functions of avatars and alt(s): A qualitative study, to present for the whole group. In last week's blog post I had pointed out that the skewed sample size (18 females and 6 males) was a weakness in the authors' method. When we discussed this in class we came to the conclusion that this might be because females are more likely play with more than one avatar (one of the questionnaire criteria). If that is the case, and females are over represented in playing with more than one avatar, than the selected sample group might be correct. If that is not the case than the sample size is skewed. 

Next we chose to discuss the qualitative method of focus groups. We came to the conclusion that focus groups are beneficial when they encourage all the participates to speak freely and be inspired by each other. However, it is not uncommon for one person to "take charge" and speak more than others, while someone might sit completely quiet. The fact that not all participants get to express their opinions is one of the main drawback of focus groups. To avoid this problem it is better to use one on one interviews, where you "ensure" that the interviewee can speak more freely without being shy or interrupted. 

After that we decided to try to evaluate the description of a focus group on the course's Wiki-page. I suggested that we should add that to conduct a focus group, you usually have to pay the participants with money, movie tickets, "fika" or something else. This is a problem because in bachelor's/master's level research there is usually no budget to pay the participants. That means that it could be hard for master's students to conduct focus groups. Another drawback is that even if you have a budget, the fact that you pay the participants could affect the outcome of the study. I think that there is reason to believe that people could behave differently if they are payed to do something compared to if they do it for free or out of pure interest. 

To conclude, I think that this was a good seminar with some interesting discussion. I found it interesting that people had read papers with such different qualitative methods, ranging from interviews to diaries and narratives.  

Theme 6: Qualitative and case study research (Pre)

I chose the research paper The psychological functions of avatars and alt(s): A qualitative study which was published in the journal Computers in Human Behavior. This journal has a current impact factor of 2.067 and a 5-year impact factor of 2-489.

The aim of this paper was to assess how people use multiple accounts in virtual worlds such as Second Life. Previous research have shown that roughly 50 % of the users have one main avatar and another, “second” avatar. The method used in this paper was a qualitative analysis of 24 semi structured interviews. To make sure that the participants had sufficient experience of Second life the authors had some criteria when recruiting them; they must have been a “resident” of Second Life for 6+ months, be at least 18 years old, speak good English, etc.

The most obvious benefit of this method is that the authors could extract in depth knowledge of how the Second Life players use their main and second (and third, in some cases) avatar. Most of the questions that were posted were open ended, thus allowing the participants to speak freely about the subject. The interviewer followed a protocol the interviews were semi-structured, meaning that follow-up questions could be asked.

One drawback of this method is the small sample size and the skewed relation between male and female participants. From a total of 24 participants 18 were female and only 6 were males. I do not know if this is the normal composition of Second Life, but it really seems as if female participants are overrepresented in this study. The same can be seen in the racial composition, where I believe Hispanics and Blacks are underrepresented compared to the American society in general (8 % and 4 % respectively).

When reading this paper it clear that the qualitative method is a great was to get a comprehensive view of a certain subject. I also like how semi structured interviews and follow-up questions could be used to gain more knowledge than you anticipated.  

A case study is a research method were you look at a specific event, phenomenon or group of people. The aim is to thoroughly examine and analyze the event within the context that it occurred. Data collection could be done through both quantitative (surveys) and qualitative (interviews) methods. The case study is most commonly used within social sciences but it could also be used within natural sciences.

The second paper 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 which also was published in the journal Computers in Human Behavior. In this paper the authors examine the event where the University of South Florida Polytechnic was restructured / closed down. The main focus was to look at how students, media and people involved used different media (social and traditional) to discuss this event that was ongoing for 12 months. Data was collected through a questionnaire that was sent out to 244 students at the affected university. The questions concerned standard demographic data, level of internet and social media usage, news consumption and how the students acquired data about the pending changes to their university.

In the paper Building Theories from Case Study Research Eisenhardt presents guidelines for successfully conducting case research. When looking at these guidelines I see some strengths and weaknesses in the study I selected. In the second paragraph Eisenhardt states that it is important to examine a specified population when working with a case, something that was clearly done in my paper when the authors chose to look at students. The third paragraph states that it is recommended to have multiple methods of data collection to strengthen the grounding of theory. This was not done in my paper, since the authors only use the quantitative method of a questionnaire. Furthermore the data collection should be flexible, which I do not think a questionnaire is. A questionnaire is very static and could not be adapted to the participants, as opposed to a semi structured interview with follow-up questions.    

References

Gilbert et al. (2013). The psychological functions of avatars and alt(s): A qualitative study. Computers in Human Behavior 32 (2014) 1–8

Eisenhardt, K. M. (1989). Building Theories from Case Study Research. Academy of Management Review, 14(4), 532-550.

Kelling et al. (2013). The tweets that killed a university: A case study investigating the use of traditional and social media in the closure of a state university. Computers in Human Behavior, Volume 29, Issue 6, November 2013, Pages 2656–2664.

Theme 5: Design Research (Post)

Last week I read the two papers Turn Your Mobile Into the Ball : Rendering Live Football Game Using Vibration by Li et al and Comics, Robotics, Fashion and Programming: outlining the concept of actDresses by Fernaeus & Jacobsson. Both these papers were related to last week’s theme design research. Before I read these papers I had my own view of how a prototype works and how a prototype is created. Since I chose human computer interaction (HCI) for as bachelor I have a hands-on experience of prototypes. I remember spending a lot of time going from an idea state to prototype state by using mockup programs such as Balsamiq and Mockup Builder. Even though the prototypes made in this week’s papers were more elaborate than my mockups, they still accomplish the important task of showing how a product or service might work when it is completed.

I also read some of the other blogs to get a deeper understanding of how other people interpreted this theme. It seems as if most people have a view that is somewhat similar to mine. Most of us are studying media technology (and some people have the same bachelor as I have – HCI) so it is natural that we all have some experiences with design research, prototypes and mockups.  

Unfortunately I was not able to attend any of this week’s lectures. I had a clash with another course during Monday’s lecture and yesterday we spent all day rehearsing our Future of Media presentation. From what I read in the other blogs Ferneaus lecture seems to have been was very rewarding, so it is too bad that I could not attend that lecture. 

Theme 5: Design Research (Pre)

The first paper I read was Comics, Robots, Fashion and Programming: outlining the concept of actDresses written by Fernaeus and Jacobsson. In this paper the authors evaluate how the fields of comics (as in comic strips) and fashion could be used in combination with programming and robotics. Some of the current consumer robotic products, such as the Roomba or the Pleo, do not have GUIs, monitors or any other usability friendly interface to interact with. Therefore the authors propose that users could interact with these devices by adding or changing the robots' clothes or cloth covers. In the case of the Pleo, a robotic household pet, a bracelet and a pajama with RFID-tags are used to intuitively change the mode of the pet from “watchdog” to “sleeping”.

I don’t know if I learned anything in particular when I read this paper, but I think it is an interesting field of research. The connection between fashion/comics and programming/robotics is very original and not something I have seen before. I also liked how the authors used pictures to clearly show how their design prototypes worked.

What role will prototypes play in research?

I believe that prototypes already do play an important part in research, and I think that they might become even more important in the future. As technology and media technology are growing more complex, it is becoming increasingly difficult for users to “imagine” what a product or a service might look like. When reading the article Turn Your Mobile into the Ball: Rendering Live Football Game Using Vibration it is very obvious that the prototype that the authors constructed played a major part in their research. Without the prototype it would have been almost impossible to conduct any usability tests. The late Steve Jobs once said that “people don't know what they want until you show it to them”, and I think this is very true in technology research and business. How are people supposed to know if they want or need a product if they have never seen the product before? The funny thing is that this view is not at all new, since Henry Ford back in the early 1900’s said that “If I had asked people what they wanted, they would have said faster horses”.

What are characteristics and limitations of prototypes?

The characteristics of a prototype could vary greatly between different areas of research or business, but something they all have in common is that there are used to test and evaluate a new design. There are several reasons why a new design needs to be tested. One reason could be usability aspects, as seen in the paper by Li et al. Another could be durability or feasibility tests, as seen in cars or larger mechanical products. One limitation of prototypes is that there are usually only produced in a small scale, and therefore the price per unit is very high. Because of this it is hard to determine the “value for money”-aspect and what the final price might be when the product reaches mass production. Another issue with prototypes is that they do not represent the final product, and therefore it could be treacherous to base the perceptions of a product on just a prototype. Even though a prototype is supposed to be quite “usable”, it could still be much worse than the final product.

References

Fernaeus, Y. & Jacobsson, M. (2009). Comics, Robots, Fashion and Programming: outlining the concept of actDresses. In Proceedings of the 3rd International Conference on Tangible and Embedded Interaction. New York: ACM.

Réhman, S., Sun, J., Liu, L., & Li, H. (2008). Turn Your Mobile Into the Ball: Rendering Live Football Game Using Vibration. IEEE Transactions on Multimedia, 10(6), 1022-1033.

http://www.goodreads.com/quotes/369912-people-don-t-know-what-they-want-until-you-show-it

http://www.goodreads.com/quotes/15297-if-i-had-asked-people-what-they-wanted-they-would

Theme 4: Quantitative research (Post)

I attended a seminar two days ago in which we discussed how quantitative studies are used in research papers. We were divided into small groups and were asked to present the papers that we had read and what kind of studies that had been performed. For some reason we all found major methodological problems in our papers, such as one study collecting Internet behavioral data from 2003. Even though the study itself was performed 2011 or 2012, the authors chose to use “old” data concerning how people discuss politics (online, offline, both, or none). We all felt that data from 2003, when the Internet still was “young”, cannot be applied to draw conclusions about the current Internet behavior. I do not remember the exact problems in the other studies, but we all agreed that they all had some methodological issues. An interesting remark is that one of our group members found a paper which used both quantitative and qualitative method at the same time. The authors of that paper conducted deep interviews with 20-something participants, but analyzed and presented the findings at statistics/data. This was interesting because I have never really heard about that method before. According to Hrastinski this is called a mixed-method.

After presenting and discussing all our articles there was unfortunately not much time left to pick an article to present for the whole group. We decided to pick my article, which was the one with the least methodological problems. In short my article evaluates if student prefer email or face-to-face communication when they need to get in contact with their teachers or faculty advisors. The authors collected data from 300 university students through a questionnaire, and it is therefore a clear example of a quantitative method. From the data the authors drew the conclusion that students prefer FtF-communication in most cases, unless it is just a simple question that could be answered quickly by email.

We tried to describe the article/method through a written flow-chart, but because of the lack of time we could not really describe the whole research process in the detail we wanted. 

Theme 4: Quantitative research (Pre)

I chose the paper Students’ Perceptions of E-Mail Interaction During Student-Professor Advising Sessions: The Pursuit of Interpersonal Goals, which was published in the Journal of Computer-Mediated Communication. The aim of this study was to assess if students prefer e-mail over face-to-face (FtF) communication for interpersonal communication with faculty advisors.  A questionnaire was sent to 300 students at a university in the northeastern United States, asking them to answer questions about how often they use email, if they prefer email over FtF, their perception of self-presentation in emails, etc. It is clear that this is a quantitative method, since there are no in depth interviews and the questionnaire is sent out to a relatively large group. 

I learned how quantitative methods could be used to not only to answer the main research question, but also to evaluate differences between sub-groups of the sample. I think one weakness in the use of the quantitative method is that the sample group might be a little too small to conduct sub-group analysis on. Even though there were 300 respondents in total, the number of African American males (which was analyzed as a sub-group) was much smaller. In order to get statistically reliable results the size of the subgroups (and therefore also the total number of respondents) might have to be increased.

The paper Physical Activity, Stress, and Self-Reported Upper Respiratory Tract Infection evaluates how the incidence of upper respiratory tract infection (cold/influenza) is related to the subjects’ level of physical activity. Data was collected from 1509 Swedish males and females in the age 20-60 years old. The results suggest that there is a strong negative correlation between the occurrence of URTI and the level of physical activity. People who report high levels of physical activity in their everyday lives have an 18 percent reduced risk of URTI compared to people with low levels of physical activity. Even though this paper is far outside the area of media technology I found it somewhat interesting to read. I like to see how “real” scientific papers are constructed in terms of methods, language, presentation of results etc. Also, something that I found interesting is that one of the authors, Mats Lekander, did the statistical analysis for my dad’s PhD thesis. I have also worked with him a couple of times at the Karolinska University Hospital in Solna.

A good example of a quantitative method would be the “partiundersökningar” / election polls that are held each month in Sweden. In these polls there are usually a thousand or two thousand people that are being asked which party they would vote for if the elections would be held today.  I think that the main benefit of a quantitative method is that it helps you to get a broad overview of a subject. If a sample group of people is a representative for the society in general, the results could be extrapolated and applied to much bigger groups (perhaps Stockholm or Sweden). Downsides of this method are you often get very “raw data” or numbers from the results. A study with thousands of people has to be constructed in a very general way so that the results can be statistically analyzed. Therefore you might miss out in important information or subtitle opinions of people. Going back to the election polls, we see that a quantitative survey will produce a statistically reliable data of which party people would vote for, but it would not explain why people vote as they do or how they perceive the different parties.

Qualitative methods on the other hand are perfect for in depth knowledge and personal interviews with subjects. Aging using the election polls as an example, a qualitative study could be personal interviews that are made after the quantitative study to collect more in depth data. Benefits of these methods are that you get a deeper understanding, and interview subjects could provide you with valuable “plain text” answers instead of just raw data. The interview does not have to follow a strict script, and the direction of the interview could be changed in real time with for instance follow up questions. Limitations of qualitative studies are that small groups of people might not yield statistically significant / reliable results. Also, it is very hard and takes a lot of time to present the finding of in depth qualitative interviews in a good visual way.