Comments

on Jakob Forell

In answer 2 you write that " we can [still] describe a proposition about an object named X as long as we know we describe a true statement of fact about the real X." Do you mean that the difference is similar to the difference between knowledge by description and knowledge by acquaintance? I.e you can propose something you've learned from some one else's description, while you can state something you know from experience? How does this relate to other verbal expressions?

answer on Matts Höglund

You are absolutely right about the relevance of the field and the fact that it is applicable to current times, but my point was that maybe we could have read something more recent and more modern? Personally, I think that reading a such contextually based and highly philosophical text is concentrating on aspects that draws more than it gains. Instead of only focusing on what is the red line throughout the text, I also had to focus on the historical aspects, which I think are partly relevant, but to what end? Was the medium for the main topic the right one for me? No, that's my critical view of this seminar readings, while the content is quite interesting.

on Johan Weinl 

Have you ever been reflecting on how you relate the two topics, theory and quantitative data? In my understanding, data can be related and relatable to a certain theory, but is not theory by itself. Let's say you've made a simple experiment on when water starts to boil. When you analyze your data, you notice that the boiling point is around 100 C and concludes this to bee your theory. This theory is also very close related to another theory that says that the boiling point of water under normal pressure conditions, is about 100 C. So the data points by themselves are not theory, but the conclusions you can draw from them is theory.  

answer on Matts Höglund

Depends really on the features and the nature of the rotting process. If rotting means that the taste and colour decreases, then maybe the system could still see the relation between the two (since both decrease), but if the colour and taste just were different, then the risk is big for a false classification.

Yes, if you add features that can characterise apples even if they are rotting, then you could get rid of these outliers. This means that what you want is data that is a good representation of the whole apple population, i.e. you wanna include rotten apples in your data. Cause rotten apples are still apples. z

on Andreas Sylvan

You write that you didn't find any of the flowcharts very useful and that "they were either unreadable or a bit confusing". I totally agree with you on this point, since I also had some problem in understanding these "flowcharts".

The problems is that (in my current understanding) they are not flowcharts, they are relation charts. I.e. how does the different variables in the research relate to each other? For example, the chart on study behaviour (https://www.kth.se/social/upload/529c40c5f2765427dc44ec15/Bild%202013-12-02%20kl.%2009.08_large.png) does not explain the flow of the method, it explains how the study behaviour can be classified into different types of disruptions, that in the research correlated with study results. 

Do you agree? At least most of them makes more sense if you look at them as relationcharts.

on Ragnar Schön

What I find interesting in your reflection is when you write about the two different kinds of quantitative research: Surveys and and measurements. In the article of Eisenhardt (see readings for theme 6), the author makes a separation between qualitative and quantitative research with the description:  "...qualitative (e.g. words) and quantitative (e.g. numbers)...". But in my understanding (as in yours), a large number of words also constitutes a quantitative research. 

So do you have any ideas on where to draw the line? When does qualitative research become quantitative, if you look at it in terms of Eisenhardts' definition?

Theme 6: Qualitative and case study research post-reflection

Personally, I have sort of been using a case study methodology but without me knowing it. Almost two years ago, in 2012, me and Axel Hammarbäck wrote our bachelor thesis together about gamification and how gamification can be applied to math education and possibly improve students understanding of mathematics. Our methodology was based on a math game that we created and tried on some secondary school pupils at a school in Stockholm. On the occasion, we evaluated their skills before and after, and also used interviews, questionaries and observations to triangulate and find out what the pupils attitude were against such an application. 

Comparing our methodology to the one that Eisenhardt describes in his article, one can see some similarities but also differences. Firstly, we used theory and other applications to build one main research questions that was to be answered with the help of three sub-questions. We didn't not have an hypothesis but just a main focus that we wanted to answer. 

After we were done with the focus, we tried to find a case or setting to investigate. We found a high-school that let us in to do our study that fur filled some criteria that our research focus demanded and went through with our study. After transcribing all the data, we started to analyse it in search for general trends and general patterns in the data. These trends and pattern together later formed our conclusions and our theory building. 

As mentioned before, I had no idea that we actually performed a case study when we wrote our thesis. For me, case studies were something that economists and law students used in everyday education, the familiarise themselves with their future work that often is case based. Reading the article of Eisenhardt, I quickly saw the similarity and understood both what was good with our metrology and what could have been improved. See, what we did not do was to search for cross-case patterns or validate through other experiments. Of course, lack of time were an issue, but the research would probably have gained a lot from doing the same study on a different setting. 

Theme 6: Qualitative and case study research pre-reflection

Qualitative Methods

I have read a paper from the journal "Media, Culture and Society" on trying to find characteristics of emerging adults who do not use social-media. This is done through identifying non-adopters from a large corpus of interviews not intended for this research (National Study of Youth and Religion, NSYR), but still including relevant information. In this corpus, 20 persons of equally varying gender are identified as not users of social-media (Facebook or Myspace) and their interviews are used in the study. The results from the deep-analysis of these interviews are then compared to four random picked social-media users groups (20 persons per group), to confirm the differentiating characteristics of the non-adopters.

This article uses the qualitative methods: Interviews and content analysis. Since the authors did not perform the interviews, I would consider the main method used to be content analysis, but since they are still interviews, the fact remains. The benefit of using these methods in this research is mainly due to the external research material. Interview design and gathering interview subject can be very time consuming and using already collected data as in this research can enable the time to be spent on a deeper analysis or other important research methods. Because most emerging adults are social-media users, there are only a few non-adopters in the data. A quantitative method would not find any statistical significance while using qualitative method on the few non-adopters can gain the research much more material to work on.

On the other hand, using an existing interview corpus can also have its limitations. Since the corpus might be designed for a totally different purpose, the information might be biased towards the design purpose. Like in this research, one therefore needs to validate the results inside the population used to prevent this bias to be incorporated in the results.

The only critic I have on the research is also what is a limitation of the research. Since they separate social-media users and non-users through one question: "Do you ever use social networking websites such as MySpace and Facebook?", social networking users that are using other social networking platforms that cannot be compared to Facebook or Myspace but that are still social networking platforms, are not taken into account. Social networking can also happen on forums, through email, dating sites etc. These two were by the time the two most popular networking platforms, but still there might be some relevance in comparing "social networking" to the two biggest ones.

Reference:

Bobkowski, Piotr, and Jessica Smith. "Social media divide: characteristics of emerging adults who do not use social network websites." Media, Culture & Society 35.6 (2013): 771-781.

Case Studies

A case study is a sort of qualitative study where the researchers concentrate on a specific case in creating theory. The case can either be created specifically for the research or an historic or past recorded event. For example, one could ones iPhone application through controlled app testing, where you aim to investigate if the interaction is good or not, or one could investigate the phenomena at the hostage taking at Norrmalmstorg (called the Norrmalmstorg rubbery), where the hostages sympathised with the hostage-takers. During the case you need to collect data in form of observations, questionnaires, interviews etc. that you later analyse in order to form some kind of theory. What is separating case studies from other studies is (inter alia) that case studies starts with a broader hypothesis than other studies. This broad hypothesis is later focused throughout the research with the help of collected data.

In media technology, it is more common to construct the cases since media research usually concerns on-going settings and situations. I have read the article "The Use of Instant Messaging in Working Relationship Development: A Case Study" where the authors try to investigate how instant messaging (IM) is used to create and how it improves or impair working relations inside or between departments and ranks. This is done through focusing on the case of a Korean tire company. The authors used interviews, surveys and social network analysis in the case study in order to triangulate, i.e. see if the same patterns cause by a specific phenomena show up in different analyse methods.

If you analyse the article in terms of the process that Eisenhardt suggests in the article, you see some similarities and some differences. It is of course difficult to follow the whole research process through the paper, since not all planning steps are recorded in the article, but some parts are mentioned. For example, the research questions are not (as Eisenhardt suggests) broad and quit focused, which is definitely a fall back compare to Eisenhardt. This results in no observations collected and therefore the research is not gaining as much as possible from the research method.

Reference:

Cho, Hee‐Kyung, Matthias Trier, and Eunhee Kim. "The use of instant messaging in working relationship development: A case study." Journal of Computer‐Mediated Communication 10.4 (2005): 00-00.

Theme 5: Design Research pre-reflection

Comics, Robot, Fashion and Programming: outlining the concept of actDress

The article is concerned with physical programming and how inspiration for such physical programming languages can be collected from comics and fashion. Reading the article, I was faced with confusion about the meaning of physical programming and did not find any definitions through searching the Internet. The word "physical" indicates that is can either be programming physical factors (such as movement) or programming by physical means in terms of clothing etc. After reading the article I understand in which sense they use the term, but I am still a little confused because of the ambiguations. A short definition in the introduction could have been helpful in getting a deeper understanding of the problem domain.
The key points of the article however concerns physical programming in terms of how symbols and clothing could improve the usage of robots, i.e. so that the purposes of the robots can be understood and easily perceived but also changed for example according to changed clothing. From observations from fashion and comics, the authors suggest a physical programming language called actDressed and explore its application it in three different scenarios.
Since the authors seek to make a system for physical programming and to maximise usability, I find that the article totally lacks evaluation methods. How can we ultimately know (as in creating theory) that this is a good system if nothing but "common semiotics" are used and not evaluated?

Turn Your Mobile Into the Ball: Rendering Live Football Game Using Vibration

The article is about trying to get users to perceive a football game through tactile feedback (vibration). The authors have designed a device that gives users feedback on information in a football game, such as goals, team ball possession and shots.
The hypothesis that the authors are trying to investigate is how vibration can be used to communicate information that can be recognised by users only through tactile feedback. So as in a user oriented research or media technology research that is concerned with ensuring that information is communicated, users somehow have to bee included to ensure that it actually works. The research in question includes users by developing a functioning prototype that is then evaluated on actual users. Without the prototype, how could one evaluate if the hypothesis is good or bad?
But of course, as the word suggests, a prototype is a prototype and not a ready-made application. I.e. prototypes are used to test the fundamental function of a device but do not resemble a ready-made product. So the step from prototype to product might produce much better results or much worse. To ensure that a prototype is not far fetched and actually focuses on the factors that one is trying to measure, some kind of theory or proof that the prototype is based on is preferable. In the research in question, the basic functioning of the prototype is based on other research and theory, in form of restrictions of the vibration-parameters and guidelines for application of these.
What I find really well performed in the paper, is the identification of the important factors of a well working interactive product: Efficiency and user satisfaction. After identifying these factors the authors also concentrate on trying to measure how these factors are furfilled in the users experience. These measures are then communicated through diagrams and other correlation measures etc. and really give you a good overview of how the prototype is performing.
All in all, a really interesting and well-made article!

Theme 4: Quantitative Research post-reflection

As mentioned in my post-reflection, this weeks topic have been very common in my studies during the last year. I mainly focused on studies in machine learning and audio content analysis, in which quantitative research is the very common. Machine learning is all about trying to make systems that learn from data in order to either predict the future and make decisions accordingly. E.g. if you want to classify apples according to their colour and acidity, you show the machine learning applications a lot of apples (in form of the measures for colour and acidity) and tells it which kind it is. After showing it a variety of apples (the more the better), the application soon will get confident enough to make its own decisions if the next apple is a royal gala or granny smith.

The machine learning workflow is in some ways really interesting to relate to theory and quantitative methods. In quantitative methods, the workflows foundation is data that is collected through experiments. This data is then analysed in search for patterns that can reveal some correlation between variables or that seem to be specially relevant to predict future outcomes. 

Then of course, one needs to somehow statistically show that the correlation is relevant and probable, which is also a very important step in the machine learning workflow. Since the machine learning applications are self driven and not human controlled, the application need to learn how to ensure that its classification is probable and not obviously false and thereby forming some kind of artificial intelligence. This can be done with such statistical measures, i.e. measures that can show correlation between data variables and outcome. For example, let's say that the apple classifier suddenly was to classify a rotten apple. The colour is different from anything it has seen before, and the acidity is also quite unusual. In this case of an "outlier", it is important for the classifier to understand that it does not understand and output this uncertainty instead of a probable false classification. 

In the same way, quantitative methods need to be based on a variety of data that clearly reflects the reality it tries to measure and also need to be ensured to have correlations between variables and result.

Theme 4: Quantitative research pre-reflection

 I have read a paper on pitch extraction from noisy signals by Tetsuya Shimamura, Member and Hajime Kobayashi, in which they try to improve the "standard" Auto Correlation Function (ACF) used to estimate the frequency, in which vocal folds oscillate when speaking. This can e.g. be used for karaoke games (such as singstar) or for analysing speech to recognise language.

reference: Shimamura, Tetsuya, and Hajime Kobayashi. "Weighted autocorrelation for pitch extraction of noisy speech." Speech and Audio Processing, IEEE Transactions on 9.7 (2001): 727-730.

In the paper the authors propose a improved estimation algorithm, which they test and evaluate on eight 10 s long speech tracks (4 male and 4 female). This makes a total of 80 s of data which is divided into parts of 23 ms each, resulting in ca. 3470 data points. All these data points have assigned true values to them and the testing is done through adding noise with different level (SNR Inf,10 dB, 5 dB, 0 dB, -5 dB) to the speech tracks and by then measuring how well the algorithm estimates the pitch for each data point and each SNR level. Then the authors calculate the absolute distance between the estimated pitch and the true pitch, and tolerates a difference on ± 10 Hz. They also compare their algorithms result to three other algorithms results in order to show the improved performance.

The method used in the paper is quite "standard" in the field and since I have been working on project in almost the same field, I didn't learn that much from the method used. On the other hand, this is a solid method for evaluating the algorithm since it evaluated the relevant factors, that are also the aim of the paper. The authors try to improve the algorithm and therefore they also compare the improved one with the other ones to ensure that they have succeeded. They also try to estimate the performance with different levels of noise, which also is exactly what they mathematically prove the improved algorithm to handle better than the other ones.

The critic I have on the method is the tolerance and the data they use. Since frequency is not linearly perceived by our ears, a 10 Hz tolerance on a estimated 200 Hz pitch is less perceivable than a tolerance of 10 Hz on a 100 Hz pitch, which is roughly the mean pitches for men respectively women. The data might also be biased, since they only use 8 japanese 10 s sentences. The human speech system works universally in the same way, but language is different. Maybe the algorithm only works on japaneese? There is no discussion about this in the paper.

Physical Activity, Stress, and Self-Reported Upper Respiratory Tract Infection by Olle Bälter et al.

The study is a population-based study performed in a middle-size county with an "normal" rate of urbanisation. The researchers sent out web questionaries by email that the subjects were asked to fill. The subject were also given follow-up-questionaries to see if their circumstances changed. The questionaries included questions on how much they do physical activity, perceived stress, age, gender and other relevant information. From these data, the researchers tried to see patterns in what factors possibly could impact how much we suffer from URTI. This is done through calculating the risk of somebody developing URTI and comparing these with different group inside the population (men, women, young old etc.). They also try to fir the data to poisson regression models, i.e. they try to predict how the data can be predicted in the future.

The quantitative method used in the paper is based on questionnaires where subjects express how they think they are. These data might well be biased by what people like to think about themselves and might not reflect what they actually are. This uncertainty can of course be generalized through using many subjects, but could still make an impact on the research. Even though using these questionnaires as a base for the analyze, the authors uses statistical methods for validating the data and seek correlations between different groups in the population. In the end they conclude that there were a connection between physical activity, stress and URTI and this is done with the help of such measures and methods.

Theme 3: Research and Theory Post-reflection

 The previous weeks seminars included interesting assignments and discussions leading to an improved understanding of what theory is and is not. Through these seminars I broadened my understanding of what the different kinds of theory (according to Gregor) mean in practise and how to classify different theory into different categories.
Interesting to see was also that the course-collective definition of theory varied between participants. Even though the text have been corrected by other course participants from different years, the text still had some sentences that was questionable and also “corrected”. This shows the relevance of the discussion of theory and its characteristics and the diversity of peoples understanding of the theory about “theory”.

So why is this important to do? In the same manner as Gregor aims to understand and categorize different theory, it is of course important for me to do the same when faced with a scientific task. Understanding what the theory aims to do will ultimately help me in understanding what I am doing when I apply certain theory to my research.

Apart from a deeper understanding of theory, I also learned to analyse text in search of theory and also critically analyse this theory in its context. By reading articles and identifying the different theory and category, one can easily analyse the relevance of the theory in the context of the paper, and also see how the theory is applied and applicable to the method and results. This is a good exercise do to improve analysing skills as ones own writing skills.