onsdag 18 december 2013

comments


To Johan Weinl
Like your ability to make an good answer short, but still give a clear answer. It’s something that I cound find difficult to do. But some consepts, like ”define defition” could be make clearer with help of an example or further explination.

To Ekaterina-Karpukhina
I like the way you described Sense-data. You include and discussion on knowledge by acquaintance and knowledge by description, in a good way. Compared to others, including myself, who tried to get a more direct explanation or definition on sense-data. It's maybe easer to get an idea of what sense-data is in those more direct description. But I like the fact that you show a relevant discussion in your answer, it's makes it clearer that you really have read the text.

To Oscar Friberg
I find your answer for question two interesting. Even if your answer could be and probably is true, it opens up for some discussions. Its true that if an experiment show same result on 100 tests, we could not say for sure that the same result will come on the 101 test. The results will thereby not show us an absolute truth. But if a large amount of testes bring the same result, the result are approaches a statement of fact. Questions that comes up when I read your answer is, is there a limit of same results needed, before we can call it statement of fact? If so, how what is it? (expressed in numbers) And could an experiment be in an grey-area and thereby be both a proposal and statement of fact on the same time?

To Jakob
I also think the authors are narrow minded, and have hard do understand there criticism. But with knowledge of the author’s background and in what time it was written make it somehow easier to understand, I think.

Do you after this week’s lecture and seminar have the same idea of what is "old", and "new" media? I think about the power the new media has, hand how it can "manipulate" people, or how it could be consider as a threat to culture and art.

To Johan Weinl
I also have a hard time to answer all the questions, by just reading the text. Even if I think you have clear answer on all questions, except on question 3. Do you have an idea of what "old" media is, thanks to the seminar or lecture? And you find any topic or concept where interesting in the seminar?

To Carl Arhsjö
I had a hard time to understand was "myth" is, but I like your description of myth and think you have a good discussion on how enlightenment is related to myth, I cant do otherwise but to agree with you. But I had a slightly different sight on what "culture industry" is. I think Adorno and Horkheimer in a way claim that it is something undemocratic, since the media was and is often owned by big companies who decide what the media should contain. I don’t think it’s the truth for today, thanks to our many media platforms, and the consumer in a way has become the producer.




Good that you find a interesting paper, I wonder from what journal? I also wonder what kind if theory do you think they use (analysis? explanation?), and even if they presented the theory in a good way, do you find anything that the authors could make better?

Hello! I wonder little what you think about your paper, was it a good/bad paper? Does it exist something that was weird or that the authors could make better?

Did you find it easy to decide which theory type that wear used in your paper? Would you argue that it exist only one correct theory type, or is it up to the reader to decide which theory type the paper belongs to? Could several types be right way to describe a theory?

To Martin Johansson
When I first read Ylva’s research paper, I thought that a user study was missing. But after the lecture, I don’t think so anymore. Because the purpose was not to create a final product that people like or want to buy, the purpose was to show that interaction with robots could be made on several ways, even with physical objects. However some things in the text will not be proven, because of the missing user study. Like we don’t know if the products became more playful with the new interactions.

On my blog


Hi Tommy! Thank you for your post, I think that you made a good compilation of our work at the seminar. You write that paper-based surveys have a higher response frequency then web-based ones. I remember that Olle Bälter said so and I was really surprised. I really do not understand how it is possible. I think that the filling of paper forms requires more action from the respondent. He must specifically find time to fill it. Besides, to handwrite is harder than to make some click on the mouse button. Moreover, most likely you want the person to send the questionnaire by mail. I do not understand why then it provides more responds. What do you think?”

Answer: I think that paper-based surveys often are often given by a physical person in front of you. In that way it is more personal, the participant feels like he/she doing a favor. And it’s very easy to hand in a paper if the person (which the case often is) is in front of you. If you think of course-evaluation for example, it’s easier for the participant if the teacher bring a paper that he/she hand out and then collect. Compared to if we students have to go to a computer and do several clicks on links in order to get to the survey. So my answer is that I think that paper-based surveys could be the easiest way for the participant.


As you pointed out sometimes theory is misinterpreted or confused with hypothesis and data. What I am curious to learn about is how the peer reviewing process actually works for journals and what steps do they use to ensure that the papers that are published are good quality?

Answer: have a weak memory about Stefan said something like papers in journal will be review twice. First will some heaver problems in the papers be corrected, second time they probably focus more about the details, if their is no bigger problem in the papers. 

So the author get will get the paper for correction at least twice, and I assume that the people that review the paper is good in there job, and used to look after some common problem.

This is only how I think the problem is, thanks to a weak memory.


Answer: Stefan said that the second seminar would be more discussion about the "what theory is". Think I had bad luck to be sick almost whole this week. The second seminar was a seminar I really wanted to be in, sense the first wasn’t so much about theory. However I think your description on the seminar is quite good, and somehow answers some questions I had.

From Stefan Etoh
Hi.

I find it a little bit difficult to follow along your argument about method 1 and 2. What was these methods and why were they used? 

I'm also curious about what image technologies they used? Do they ever say why they used this type over another?

Answer: Well, as far I understand by the meaning of "method", it's how they use the detectors, sense a detector can be used i several difference ways. It's really hard to tell exactly what the different methods means, sense the text refers the explanations of the methods to references. For one method, they explain quickly with quite hard math formulas (that I don't understand), and write "this formula combining with (reference)" do that. So to be honest, even I don't really know the differences and therefore I argued for the text is of an analysis theory type.

Almost the same thing with the technologies, but they mention the benefits with the different technologies, and say like "this is better to detect the pattern of couples". To really understand the technical part, I think I also need to read the text references.

From Zahra A
Hi.
It seems like you chose a really confusing and difficult paper, well done on trying to explain and make sense of it anyway!
I'm wondering though about the proposed new strategy "that makes use of the pattern that appears when people overlapping each other". If I understood correctly, this is the 2nd method, right? Did the author explain the process of how they came up with that new strategy, or is that also only referenced to? By your description, it seems like they briefly proposed the strategy and then went right ahead with the study of the two methods to answer their questions about the detectors.

Answer: The way they make us of the pattern that appears when people occluded each other is not the 2nd method, it’s the second detector. As far as I understand, there is different kind of methods to use the detectors and they try two different methods when they text compare the new "joint detector" (that make use of the pattern of two occluded), and the old single person detector. However they never use the first method for the joint detector while they use both the first and second method for the older "single-person detector"


What I learned, week 51


This week has been about qualitative methods and about case study research. I think both of my papers give me some new experience when it comes to methods.

The paper I find that using qualitative paper (Lomanowska & Guitton, 2014) was about  find out more about the representation of pregnancy, birth and maternity in virtual environments, and the players behavior and opinion in this context.  The study was made in the game Second life, and the methods that were used were observation and analyzing. They was observing the game for nine months, the first six the month was to identify the main themes related to the activities in related Sims (avatars).  Related Sims was the Sims that has the word “pregnancy” or “maternity” as keyword. They used qualitative analysis of online material, in order to answer the “5Ws and H”, the question of: “who, what, when, where, why and how”.  The online content was 5 personal blogs, 3 journalistic reports, 2 interviews and one instructional blog.

Something that I find interesting in the methods was that the authors never made up their own questions for the research they only used online content. But after observing the game they made up some themes and analyzed of often the themed appeared. For
Example: to answer the question why? The made up the themes “Curiosity”, ”Progression of social aspects of virtual experience”, ”Emotional reasons related to real life needs and desires”, ”Sexual basis” and check how often the theme appeared in the media that they analyzed. I think it could be smart to use online content, because of the time saving. However I think this method is little weird, sense the media may not be complete, and bring all the reasons of why or one online content could have different approach compared to the others media. I don’t feel sure that the method gives any good or valid data at all.

The thing I learned most about this week is about case studies. Before this week I had almost heard of it when it’s related to HCI or market research. The case studies I read before has always been about testing some product in some way, or in a context, and then evaluate the users thoughts. So I thought that case studies was like a method to test something in some context and therefore required a user study as well. However I don’t think so anymore. In the I read this week (Fisher, Badam, & Elmqvist, 2014) they focused on the practical software engineering challenges inherent with building DUI (Distributed user interfaces) applications. They doing it by exploring three cases and with help of literature they come up with some challenges in the cases. To bring some solution they implemented three DUI applications, described in the cases and with some help of literature and by themselves they try bring some solution.  They don’t focus on things like interaction, and therefore don’t have a user study. I think they did a good research and discussed things like limitations, design guidelines and future extensions to the framework in a good way. So now I don’t see case study as a method anymore. Case study is when research uses one or some special cases to help for get some questions answered.  It’s several methods that is good compliment with case study, like analyzing, observation, dairies or interviews.



Fisher, E. R., Badam, S. K., & Elmqvist, N. (2014). Designing peer-to-peer distributed user interfaces: Case studies on building distributed applications. International Journal of Human-Computer Studies, 72(1), 100–110. doi:http://dx.doi.org/10.1016/j.ijhcs.2013.08.011
Lomanowska, A. M., & Guitton, M. J. (2014). My avatar is pregnant! Representation of pregnancy, birth, and maternity in a virtual world. Computers in Human Behavior, 31(0), 322–331. doi:http://dx.doi.org/10.1016/j.chb.2013.10.058

fredag 13 december 2013

Sem 6



The paper I find use both qualitative methods and some quantitative method.(Lomanowska & Guitton, 2014)

1.Which qualitative method or methods are used in the paper? Which are the benefits and limitations of using these methods?

They mainly used observation as method. But also did some qualitative analyses on online material, like videos, blogs, interviews and journalistic report. Observations could help the author to get some depth in the research, and also get some understanding in things the author is not familiar with. It could be helpful to see what is really happening and what questions could be made from the observation. However, it could be hard to observe everything, and things could be missed. Observations alone don’t answer all the important questions, like the question “why?” Like in the study I read, when the author do observations on player for an online computer game. If they a player do something, it could be hard to say why the player did as the player did. Even how the player did it isn’t obvious. So all questions could be hard to answer with just observation.

The paper I read use analyzing to answer the “5Ws and H”, the questions of “Who, what, when, where, way and how”. Analyzing could be good if there exist something that is good for the study. However it could be sometimes be hard to find answer on a specific question and the author can’t know for sure if it is an extreme case or general case that being analyzed.

2.What did you learn about qualitative methods from reading the paper?

 I’m not sure however I learned something new or not when I read the text. However, I think that many times quantity methods could be a good complement to quality methods, and also I think the text show that and used the both types of methods in good way.

Something that could be consider as new, is the fact that they didn’t get their own inputs with there own interviews, focus group or something similar. They just observed and analyzed things that already exist on the Internet. I have never had that idea of studding some interview that someone else have done and posted online. I like to make my on questionnaire, and get my “own inputs”. However, if you find something that you think is good and answers your question, why spend time on do the same thing someone else already have done?

3.Which are the main methodological problems of the study? How could the use of the qualitative method or methods have been improved?

I think they used the methods in a good way and don’t see any problem in the method. Maybe the fact that they did not made there own interviews, and only observed how thing work and analyzed things that already done. They don’t even ask their own questions with the quantity method, where they analyzed several video dairies. When a behavior is explode, usually the author make up some specific questions that they bring to the participants. However it didn’t seem to be necessary to answer all of the author’s questions.

1.Briefly explain to a first year university student what a case study is.

Case study is a study to observe how something works in practical. Could be some technical device in a new context, or how well something really works. The user could be asked to speak out the thought while he/she uses the devise. If the author can’t be present while the participant uses the device, the participant could write diaries about the usage. After the test period with the device, an interview or similar with the participant is a good Idea, to understand the participants thoughts. It’s not always that it exist participants, but cases could be study with related work as well.

2.Use the "Process of Building Theory from Case Study Research" (Eisenhardt, summarized in Table 1) to analyze the strengths and weaknesses of your selected paper.

My paper is pretty close to a design study, and with help of related works explores some problems and tries to give solutions on the concept of DUI (Fisher, Badam, & Elmqvist, 2014). The have cases of DUI that they explore and there emphasis was on thepractical software engineering challenges inherent with building DUI applications”. So they don’t have things like explore things like interaction, they don’t really have participants (maybe them self) and explore problem in things like  ”data transfer”, limited resources. So they don’t have any specified population, and I think they only have one way to collect data (analyzing previous and related work). However, they make good use of the analyzing and argue well. Think they have most of the things from table one, except things like “specified population, several methods”.

References:
Fisher, E. R., Badam, S. K., & Elmqvist, N. (2014). Designing peer-to-peer distributed user interfaces: Case studies on building distributed applications. International Journal of Human-Computer Studies, 72(1), 100–110. doi:http://dx.doi.org/10.1016/j.ijhcs.2013.08.011
Lomanowska, A. M., & Guitton, M. J. (2014). My avatar is pregnant! Representation of pregnancy, birth, and maternity in a virtual world. Computers in Human Behavior, 31(0), 322–331. doi:http://dx.doi.org/10.1016/j.chb.2013.10.058


onsdag 11 december 2013

What I learned, week 50


I think this weeks lectures were good, and sometimes a bit difficult. I liked Ylva’s lecture, even if there wasn’t so good responses from the students, when she ask things. But I share hers view on what research is, and she manage to break it down to its easiest form. I like the slide with her five steps of research, which said what research is all about. The five steps were thinking and search for something new, check it up, take notices, convince others, they accept. . It was much about “what needs to get done”, and not about “how to do it”. Which I liked sense the choice of the methods is depending on the topic, and what the purpose is to show.

When I first read Ylva’s study, I thought that a user study was missing, and the research would be more complete if there was a user study in the study. However, her goal was not to show a product that some would like and buy, the goal was to show that interaction with robots could be made on several ways, and even by physical object. I think she manage to do it in a good way, and after the lecture I am convinced that the research wouldn’t be more complete if it had a user study in it.

I think the lecture with Haibo Li was fun, interesting and somehow stressful. He had some good points about sometimes it isn’t always about solve a specific problem, but that the problem can may be more easily solved by look at a problem in different perspective, and try several approaches. The important part is to remember what the basic problem is.

I liked the way he defined what a great-, big-, or innovative idea is. He claimed that is how much money the idea can generate. I also like to see thing with an economic lens, however I could argue for that an idea could be good, even if it don’t generate money. In he definition on a good idea, he mentions factors like; how many people could and will use this? Is the time right? Could we do this idea to reality in a good way? But when an idea is defined like a business plan, I think ideas like those that are for goodwill (to help people) are somehow forgotten. He maybe talks about ”business-idea”, but when I was on the lecture I understand it as a idea in general.

I think the lecture was stressful when he talked about the roll of an engineer. Like the example when he argued for that an engineer must be able to estimate some measures and questions quite quickly. That math is important for an engineer. Even that we should know about 1-way ANOVA and 2-ways ANOVA is. I never heard about the name ANOVA, even if I knew what is was. But it has been a time since I read a math-course, which made me unsure on my own math skills. So the part where he talked about the role of an engineer was a stressful. However, I think that the math could come back quickly, if I start study problem that requires it again. I also think that much of the knowledge that I need for my role as an engineer comes when I start to work, and are finished with my education.

torsdag 5 december 2013

sem 5




I think the paper written by Ylva Ferneaus and her colleagues was quite interesting. The interaction with artifacts must not always be the most intuitive way; bottoms must not always be the control to the artifacts. Sometimes an artifact could be more playful, fun and perhaps even more user friendly by different kind of interactions. I would like to see some kind of study to user experience on the kind of prototypes that is suggested in the paper. To see if a user in practice think that the artifact are fun and practical, or if its boring or to time consuming etc.

Question for seminar, I think it was interesting that Ylva and her colleagues get inspiration to the interaction from fashion and comic. What other fields could it be interesting to get inspiration from, and why those fields?

Some reflections on questions:

How can media technologies be evaluated?

I think the paper written by Haibo Li with his colleagues was evaluating media technology in a good way. Media is much about the interaction between a human and devices (often technical), and it is therefore important to evaluate the experience that the users have with the devices. HCI is an important part of media technology and I think the paper evaluate things like effectiveness, efficiency and user satisfaction in a good way.

What role will prototypes play in research?

Prototypes could have a big role and is good in order to evaluate things and ideas. I think for example the idea to use vibration to render live football was really interesting. However it is a good thing to explore other people thinks it is a good idea, and if so, how it should work. It expansive to develop a complete product, and it could also be very expensive if the complete product most change a lot, because the user didn’t like it. But create a prototype much easier and cheaper to develop than a complete product. Even if the prototype is not the same thing as the complete product, evaluations on the prototype could and will often bring data that could be important to the development of the final product.

What are the characteristics and limitation of prototypes?

Prototypes are made for somehow show an idea of a product, how it will look like or how it will work. The meaning of prototype is to evaluate the product under development, in a cheap way, but still in a good way to get data that is valid for the final product. The prototype could however be very easy. Like if a computer program is under development, the prototype is not necessary going to be a look-a-like program. It could be papers with illustrations that could give the test-user an idea of how the program will work. The test user could react on thing that the user think is smart, or weird and the developer could use the data in the development of the real product. However, a prototype is not the real product, and some data for the final product could be hard to get. It will be for example is hard to do the study that Haibo Li and his colleagues did with only papers. It would be hard for a user to imagine the felling on how the touch would work and therefore it’s hard for them to give a good response.

How can design research be communicated/presented?

Once again I think Haibo Li and his colleagues did a pretty good job in presentation of theirs design research. First by using references and theories they come up with a design that could work for the product. They build a quite advanced prototype of the product, which could be evaluated by some test user. They present all the things they evaluated by surveys. They in a clear way showed the data they obtained, and correlations between different data. They even try to explain some reasons why the result was as it was, and emanation some factors that the data could depend on. However I think they only use quantitative methods, but it would also be interesting to see some result from a qualitative method, for example an interview with a user that was interested in football and an interviews with someone that where not so interested.

What I learned, week 49


Reflections on quantitative methods


From the paper I read for this week I learned that is important that the result is shown in a good and clear way, with good explanations. I think my paper had some mistakes, and even if it’s probably careless mistakes (like the sum in percentage for all men and women that was participated doesn’t sum up to 100%, but rather 98%) it lowers the credibility of the paper. It’s important to show a result that is correct, and don’t make any assumptions from the statistic, that the statistics doesn’t show (I don’t think my paper had this mistake). Like if a class get a lower grade in a specific course one year, we cannot say for sure that the student’s of that year was lazier than usually, the lower grade could also be thanks to the exam was harder than usual. It’s something impossible to know, by just looking on the student’s grades.

The seminar this week fun and it’s obtained a lot of good discussions. Sometimes it was problem in the discussions because it could be hard to discuss quantitative methods as a subject and make generalizations that holds for all quantitative methods, when a proposal of advantages came on something that only hold for some of the quantitative methods. Like if it’s cheaper, compared to qualitative methods or not. It’s hard to say, sense it’s depending on the exact method. But some conclusion could be made for all the quantitative methods, like it is a good way to obtain much data, easier to do on a big geometric area and it’s good for statistic analyzing etc. Compared o qualitative data.

Some methods can be doing on different ways, for example; a survey could be hand out by paper or by the web. Web-based surveys have some advantages and disadvantages compared to the paper-based surveys. It was advantages like it could faster be send for long distances, if it’s many participants (threshold about 3000 persons) it is cheaper, it could be logical and skip some unnecessary questions about smoking if someone already have pressed “no” for the question “do you smoke?” for example. It could also have other medias like video and audio as complement to the survey. Papers on the other hand don’t have the same risk for technological problems as the web has, and often also have a higher response frequency. We hand some discussion in the seminar on web contra paper on what was best for the environment, but don’t came up with a conclusion. I have to see some calculations to say what is the most suitable choice for the environment.

In the seminar we also discussed how important it is to test the questions if someone is going to create a questionnaire survey before handing it out to the target group. It could easily be wrong and it’s often difficult to make sure that the participants understand the questions in the right way. The leader of the seminar showed some questions that at first sight could look like good questions, but somehow easily could be understand on several ways. It was useful to see, and a good idea would be to read some more about questionnaire survey and how to test it before next time I do one.

torsdag 28 november 2013

Sem 4



Which quantitative method or methods are used in the paper? Which are the benefits and limitations of using these methods?

The paper I have read this week is called “Prevalence and determinants of Internet addiction among adolescents”(Adiele & Olatokun 2014), and it used quantitative data to examine their hypotheses. After going thru others theory and claimed that they did some research design to create surveys they made a survey-based study. Thiers surveys contain an Internet addiction test (IAT) and “EPQR-S lie scale”. The Internet addiction test is a test developed from previous studies, and “EPQR-S lie scale” is a way to see if the participants are potential liars. EPQR-S lie scale contains “yes an no” questions that most of the people will answer “yes” on, if they tell the truth. Its questions like “have you ever made your parents disappointed?” or “have you ever cheated in a game?” If a participant says “no” on too many questions, that participant will be seems as a potential liar, and the survey from that participant will be claimed as invalid. Could be a smart thing, but I don’t know for sure how well it works.

It could be good to have the kind of test as previous studies, because its more easy to compare the results from other studies, it is even possible to merge different studies to a big study, if the participants from different studies get the same question. Limitation with surveys is that it hard to get details from the participants, and thing that happens unconsciously.

What did you learn about quantitative methods from reading the paper? 

The opportunity to find out if someone is a potential liar on a survey is something I never heard of before. Even if I don’t know how well it works, it’s something that I find really interesting and something worth to look further into. I do think it’s common that people try to avoid the truth on surveys, even if the participants are anonymous. If a participants is example shamed about something, I don’t think everybody tell that on a survey.

This is also the first time I read a paper where the authors take a test that already exist, and not developed a new one for there own purposes (even if they modify some questions little for more clarification). It could, as I said be a smart think to do, if the test is good enough for the current study, because the possibility to compare or merge it with other studies.

Which are the main methodological problems of the study? How could the use of the quantitative method or methods have been improved?

I think the paper in some way failed in use of the quantitative method, because it’s important to show all calculations right, and data in an understandable way. My biggest problem is that I claimed that some calculation is wrong (maybe careless mistakes) witch make the study less credibility.  Like the meaning “ left the current study with a final sample of 450 adolescents of both sexes (47.42% males and 50.17% females).” Makes me we wonder what the rest 2.41% are, shouldn’t males plus females be summed up to 100%

Another meaning is “However, females 950%)”, they maybe mean “(50%)” but I don’t know for sure. They also have bigger error, that I don’t have place for write about here, but thanks to the bad calculations and the contradictions in the paper I think the whole paper lack of credibility. However, when statistics is the result from a study, I think it’s very important to make sure that everything is shown in a good and correct way,

Short about the text “Physical Activity, Stress, and Self-Reported Upper Respiratory Tract Infection”.

I think it was a good text that structured data in a clear way. I thought about how people (specially men) were less stressed when there was a high physical activity (many MET-hour/day). For me it’s true, but when I don’t have time for training, I have probably much in school, work etc. and therefore stressed. When I find more free time, I will train and be less stressed, most because I have less to do from school and work.

Which are the benefits and limitations of using quantitative methods?

Lot of data is obtained, and it’s easer to make generalizations. It’s a good way to get lots of data in short time, on things that people could easily answer in a survey for example.   However details like things that people do unconsciously are hard to catch.

Which are the benefits and limitations of using qualitative methods?

Benefits are that details are easier to detect. Here you can get people’s opinions as well, in focus grope or something similar. Another method is case study, which could have more observations than a survey ever will have. However, it’s time consuming and therefore difficult to do it many times, and hard to do generalizations.

Adiele, I. & Olatokun, W., 2014. Prevalence and determinants of Internet addiction among adolescents. Computers in Human Behavior, 31(0), pp.100–110. Available at: http://www.sciencedirect.com/science/article/pii/S0747563213003786.