Comments

On http://theoryandmethodformediatechnology.blogspot.se/2013/12/theme-6-qualitative-and-case-study.html

Ragnar Schön18 december 2013 17:19

Wow, a lot of comments on this one!
It is pointed out that there may be problems with bullying and that there should be some follow up as I wrote. Also, it is said that focus groups might give better results than interviews in some cases.
As always, the result you'll get from any method will rely very much on the context. To get a more complete result I think you should combine a lot of methods, and the settings within a method could also be varied greatly. One could for instance have focus groups of very varying sizes and see how that affects how participants express their views. If somebody bullies others in a focus group it could mean that they have very strong views on the subject or that they are just mean. Varying settings could discriminate between that. For the follow-up I think that the more you have the better. Perhaps one focus group is not enough.
I think the only thing that limits this is time or money. So more time and money to researchers will give better research... ;)

On http://edvardtmm13.blogspot.se/2013/12/theme-6-pre-reflection.html

Ragnar SchönDecember 14, 2013 at 7:52 AM

Hej Edvard! I like your simplistic explanation of the case study. But I think it is important to point out which context we are trying to explore. This is simply because everything is always in the context of something else but also that everything is in a mutlitude of contexts.

On http://mathoglu.blogspot.se/2013/12/theme-4-quantitative-research-post.html

Ragnar Schön6 december 2013 06:16

Hi! Interesting post. How do you think that systems should handle outliers in general? Following your example of a rotten apple, you classify it as an outlier whereas it still belongs to some variety of apple. Would you argue that it is a shortcoming of the system and that it could be fixed by adding more features in addition to colour and acidity?

On http://mathoglu.blogspot.se/2013/11/theme-3-quantitative-research-pre.html

Ragnar Schön29 november 2013 02:50

Are you regarding the 3470 data points or the eight persons as the population of this study? I find it quite hard to decide on which is "right".

I think I would have the same critique regarding the use of a single language. However, unless they have a more robust technique ready it is advisable to start in a simple manner and then add more complexity. One may argue that this is something they should address in their discussion.

Regarding the linear approach to frequency, I think it is more advisable to use percent or cent. A semitone is ~5.9% or 100 cent across the entire spectrum.

On http://theoryandmethodformediatechnology.blogspot.se/2013/11/theme-2-critical-media-studies-pre.html

Ragnar Schön24 november 2013 12:34

Of course, intelligent design doesn't meet the standards of being scientific. But if you are a believer of ID or any other type of creationism, it doesn't matter. All those fossils and whatnot could of course have been put there because The Creator felt like it or to test our faith. I mean, you could say that evolution and big bang are true but that it's some divine entity's grand plan. Scientific critique doesn't work on pseudo-science.

The idea presented in Guess' quote is quite interesting. If I understand it correctly it is basically that scientists should mind their own business and not meddle with the "transcendent". I guess this is just because this is a threat to mainly religious people's beliefs. In that sense I guess you were better off in the 16th century, when science and religion were still united.

On http://dm2572-cederman.blogspot.se/2013/11/theme-1-post-reflections.html

Ragnar Schön13 november 2013 13:54
I think it would have made more sense reading a hundred year old book if we would have had the lecture. We only got two views as of now: Russell and Plato (2400 years ago!). To get a broader understanding of the theory of science we'll need a couple more. Philosophy is, just like most other sciences, ever-evolving. It definitely didn't stop with Russell. For some other views of theory of science you might be interested in Popper, Kuhn, Lakatos and/or Feyerabend (epistemological anarchism sounds pretty cool, doesn't it? :D ).

On http://dm2572sylvan.blogspot.se/2013/11/post-theme-1-theory-of-science.html

Ragnar Schön13 november 2013 08:17

The idea that this reality is not the "real" reality seems to pop up over and over again. We find it in Platon's theory of forms and in the hindu veil of illusion, Māyā, that obscures reality. Several esoteric and occult movements have also tried to break free and to experience a more true form of knowledge (Sophia).

The notion of different levels of reality has also been explored in several, mostly sci-fi, films (e.g. Total Recall, The Matrix, Inception, etc.).

But just because there is no real truth doesn't mean nothing exists. And what does it really matter. Isn't our experienced "reality" the most real thing that we'll get?

Start the reactor!

On http://ahrsjo-tmm13.blogspot.se/2013/11/theme-1-theory-of-science-pre-reflection_8.html

Ragnar Schön13 november 2013 07:13

Hej Carl! I think you summarised the theme very good with your text. However, I would like to question your view on how statements of facts and propositions differ from other verbal expressions. You write that the difference is that they in some sense address reality. Aren't there a lot of other verbal expressions that also address reality? Isn't our entire language referring to reality in some sense or is it only like in Platon's cave metaphor that it's just a mere "shadow" of reality?

One may argue that language is everything. If so wouldn't say colour exist without a word (or another semantic referrent) for it? Some languages have the same word for green and blue. Does that mean that it is the same colour?

On http://theoryandmethodformediatechnology.blogspot.se/2013/11/theme-1-theory-of-science.html

Ragnar Schön13 november 2013 06:54

It is interesting that you point out that I missed out on the differences from other verbal expressions. I kind of got caught up in discussing the arbitrariness of truths. I don't really see how they are different at all. They are verbal expressions just as any other. Both expressions can be true in some contexts and others not.

On http://dm2572-martin.blogspot.se/2013/11/theme-1-theory-of-science.html

Ragnar Schön9 november 2013 23:32

So, do you think the apple exists or not? What is the significance of its eventual existence for science? It might be that the purpose of this reasoning is that Russell does it to encourage us to be critical (and to have a systematic doubt). If we can't even be sure that the apple exists, we definitely can't be sure about the sense-data that we can measure from the apple.

Surely, we might argue that it is irrelevant whether the apple exists and that it is only our experience of it and how we describe it that is relevant. This is what I believe, but I think that questioning reality might be a good tool to understand how arbitrary a lot of descriptions of reality are. The apple exists, but there is not only one true description of it.

On http://mathoglu.blogspot.se/2013/11/blog-post.html

Ragnar Schön6 november 2013 08:04

You write that a proposition is something that could eventually turn into a statement of a fact as e.g. a researcher has found that it is true. But isn't also possible that statements of fact could be false? What I mean is that our imaginary researcher actually believes the statement to be a fact but that s/he never can be completely sure. Perhaps this is just a part of a reasonable doubt or Descartes' systematic doubt of which Russell speaks.

What I'm cooking up here is the question whether it is interesting to speak about true or false at all. Shouldn't we focus on for example what is good for the scientific project as a whole?

Theme 6: Qualitative and Case Study Research Post-reflection

So for this week I started to look for quantitative and case-study research in some of the Sound and Music Computing related journals (e.g. Journal of the Acoustical Society of America, IEEE Transactions on Audio, Speech and Language Processing). But it proved very hard to find this kind of research in this field. During the exercise  it was said that qualitative methods was less popular in this more engineering oriented field. In "regular" Media Technology it is more popular as it incorporates more humanistic and social sciences. Still, Sound and Music Computing has a lot of incorporation of e.g. music psychology so I expected that there would be at least some papers. It might be because I only searched for the title. My gut feeling is that people don't tend to put their method in the title and that might be a possible explanation.

As I see it there is no distinct boundary between qualitative and quantitative research. Both kinds have their pros and cons and may be combined or compared in many ways. A researcher should try to find suitable methods for their research without any prejudice. The most relevant method(s) for some research may be more or less suitable, but never predetermined. When I start doing my master thesis I will try to explore different methods and find which ones are suitable. As I most likely will do my thesis within the Sound and Music Computing field I will find that almost all previous research is quantitative (or maybe not) but this shouldn't hinder me from choosing another method. However, one might argue that in order to be taken seriously within a field one should use the canonical methods. Perhaps combining methods is the way to get away with using qualitative methods in a quantitative-dominated field or vice-versa.

Theme 6: Qualitative and Case Study Research Pre-reflection

I.
For the first part I chose a paper that study first hand accounts of digital game experiences. It's empiricism is based on interviews in six focus groups with three to five participants in each (N=19). They argue that the low number of participants allowed to go deeper in their interviews (to be more qualitative). The focus groups were varied to create some diversity between them in respect to age, gender, gaming frequency, occupation etc. The 90 minutes were divided into the following: an introduction in which the moderators presented themselves and stated the goals of the focus groups, an individual task, a group discussion and finally a group task. The participants received a minor monetary reward. The data collected during the interviews is used as a basis for a discussion in which they present nine dimensions that can be used to categorise gaming experience. They argue that these dimensions allows and simplifies further research in the field of digital gaming.

Focus groups are more time-effective than solo interviews and promote a discussion. There may however be some things that participants leave out in the company of others. Perhaps focus group discussions can be expanded with a short open questionnaire afterwards asking if there was something that they'd like ot add or something they didn't get the chance to say. Also, this could be made iterative. The conclusions that the researchers make could be presented in new (or the same) focus groups in order to evaluate them. Perhaps the participants wouldn't agree with the conclusions the researchers have made if they were presented to them. There is always a subjective dimension when analysing or choosing how to analyse data.

Reference: Poels, K., de Kort, Y., & Ijsselsteijn, W. (2012). Identification and Categorization of Digital Game Experiences: A Qualitative Study Integrating Theoretical Insights and Player Perspectives. Westminster Papers in Communication and Culture, Special Edition: Encountering the Real-Virtuality: Digital Games in Media, Culture, and Society, 9(1), 107-30.

II.
A case study focus on a single setting. This setting may be a person, a group, an event or an artifact inter alia. The central thing is that with a smaller scope the research may become more comprehensive for that case. Several methods, often but not necessarily qualitative, may be used within a case study such as interviews, observation etc. It is important that the case chosen is relevant for the research that it is chosen for. The case should be representative of the population the study strives to explore. From the results of the case study a hypothesis for this population may be constructed.

The article I have chosen (Marshall 2004) tries to make conclusions regarding the effects of piracy in the music industry by looking at a specific type of piracy: bootlegging (unofficial recordings of live music that are otherwise unavailable). They mostly look at what has been said in literature and news and combine this with the views of bootleggers, collectors and music industry representatives expressed in interviews with the author. It explores a number of different facets on bootlegging including economical incentives for and against.

The strength of this paper is that it gives a comprehensive overview on bootlegging and its effects on the music industry. However, it concludes that it is not possible to generalise the case to the population. The case of bootlegging is shown to be a special case which doesn't represent piracy as a whole. The table in Eisenhardt (1989) shows an ideal case study. I think this paper lacks some of the activities of a model case study but still manages to explore the case in a comprehensive way.

References:
Eisenhardt, K. M. (1989). Building Theories from Case Study Research. Academy of Management Review, 14(4), 532-550.
Marshall, L. (2004). The effects of piracy upon the music industry: A case study of bootlegging. Media, Culture & Society, 26(2), 163-181.

Theme 5: Design Research Post-reflection

During Ylva Fernaeus lecture she raised this question: What made this project into research? I answered that the fact that it had been presented in a research paper and on a scientific conference made it into research. This might sound like avoiding the question but I think it is the proper view on research (and science). During the seminar with Leif Dahlberg we discussed what was and wasn't art. One common criterion for art states that anything is art that is presented within the white walls of the gallery, or in a more general sense, anything that is presented within an art context. I believe that this kind of pragmatic criterion might be used for science as well: research is anything presented within a research context. In a more specific sense a research paper can be considered to be research if it meets the criteria of some research journal or conference.

During Haibo Li's lecture he talked about ANOVA. This made me think about what I am missing in this course: actual methods that we may use in our future research. I know these kinds of statistical methods by their names and that they are commonly used but I do not know how to use them. It would be really interesting to learn how to go about when you want to find out whether your findings have significance or not. What are the criteria? In the statistics course we learned how statistics work in theory but not how to apply them in reality. I guess I will have to read about this on my own...

Theme 5: Design Research Pre-reflection

On Fernaeus and Jacobsson.

Their work is based on the notion that people  personalise their electronic equipment by dressing them up, applying stickers on them etc. Fernaeus and Jacobsson proposes that this may be used in physical interfaces to control and program them by physical means.

Based on semiotic theory Fernaeus and Jacobsson introduce the concept of signifier and signified. A symbol may be a signifier for different signifieds in different contexts. In this case the physical markers are considered signifiers. The action or behaviour the physical marker incites is the signified.

For their proposed design, Fernaues and Jacobsson uses symbols known from fashion and comics. They argue that fashion is used for more than clothing, it may for instance signify group belonging, attitude etc. Comics use simple and clear shapes and typography to convey affects and information.

The proposed design, actDresses, are described as physical markers that may be attached to a digital artefact. By the use RFID the artefact may detect what the user has attached to it and run the appropriate programming.

On Prototypes.

The construction of prototypes is central in an iterative design process. Early proto-prototypes, mock-ups, can indicate problems and benefits early on so that these may be adjusted. Evaluation and adjustment of prototypes should ideally be looped until there are no more improvements, at which point you are near the final design. The evaluation process may measure efficiency, effectiveness and satisfaction by user tests (Li et al 2008). 

In the case of proof of concept prototypes it is not necessarily the idea of the final product that is realised but perhaps only some aspect of it. A proof of concept prototype is useful when a designer wants to control that some aspect of an idea will work in reality. Based on how this prototype turns out there may be implications for the design choices that are to be made.

A final prototype may be relatively close to the final design, but usually has some differences. These may for example include different materials and means of production. This may lead to problems if the designers aren't observant. I think that the remedy for these problems is to prolong the iterative design process and try to minimise the distance between prototype and product.

Theme 4: Quantitative Research Post-reflection

On Quantitative studies.

It seems to me that in the field of natural sciences and technology, quantitative methods are the ones that are traditionally more popular. The focus is on easily testable and repeatable experiments from which knowledge may be deduced or induced. A simple example: a ball is dropped a thousand times and we conclude that balls fall to the ground if they are dropped. It doesn't matter if it is the first or the thousand-and-first ball that doesn't fall, both will falsify the hypothesis. However, it is only the one were we try a thousand times that is considered scientific and it is the larger quantity that makes it scientific.

As far as I know, based on the articles I have read, there are two kinds of quantitative research: surveys and measurements. Surveys are subjective opinions pushed into a template. In this way, subjective opinions are made quantifiable and as we reach a high number of participants it becomes less and less subjective. This assumes that the participants of the survey are a cross-section of the population that it strives to investigate. If it isn't we are obviously investigating something else so this is very important when recruiting people for your survey if you strive to make conclusions about that population.

Measurements are seen as less subjective but of course have the same challenges. Everything is based on what datapoints are included in your dataset. There is always a selection and often also exclusion of outliers etc. Datasets may easily be pushed into similar templates that better match your research with the risk of losing scientific relevance and significance.