22. august 2012: Lecture #1 (What is research)

What is research?
Wonder (define a problem) → Think of an hypothesis (what will happen) → How to find out? (the method/experiment) → do it (the method) → analyze it → Repeat the proces

• Isolate and define the research problem
  
• Review what others have done before
  
• Divide problem in sub problems if possible
  
• Postulate a hypothesis (likely solutions for the (sub) problems
  
• Describe limitations (things you do and do not do)
  
• State your assumptions and define terms
  
• Identify a suitable design and methodology
  
• Identify assumptions taken
  
• Collect data
  
• Analyze data, possible answer of the question and often new questions arrise.
  
• Isolate and define new research problems...
  

The book "Planning and reseach" has a very strict definition of research, focusing only on empirical research (page 5), and does not consider collection of data without any analysis actual research. The essence could be summed up by saying research is the process of gathering and analyzing data to support the researchers hypothesis, in a structured logical way. The things that make it structured and logical is the fact that all aspects of the situation is well defined, so that other logical minds can look at the results and get to the same conclusion.

Tools for research
The books mentions the importance of distinguishing between the tools used for and the scientific method itself. Tools mentioned are:

1. Collected sources of information: Like old fashion libraries or digital databases. They are important before the "true" research is begun in getting up to speed in the subject at hand. What have others already done? What new questions did they come up with? ..and in getting inspired for own hypothesis by finding patterns in historical data no one else has discovered.
   
2. Computer technology is mentioned as a tool, and is so in the sense it can help with storing, searching and sorting data if you are able to guide it in doing so.
   
3. Measurement: A measurement can be about anything, and different tools and methodologies are invented to observe difficult phenomena. It could be telescopes and microscopes (both within and outside the visible spectrum), weight, length, different kinds of forces, direct or indirect observations. One way of indirect observation might be finding something that is correlated with the item you want to study and then look for changes in the correlated item.
   
4. Statistics are methods applied on data from measurements to guide us in understanding them. Often based on probabilistic calculations and finding trends in data (correlation).
   
5. Language is the basis of abstract thought, and in a sense language is a way of packing lots of information in terms which can be combined to make up more abstract terms. This is why it is so important to learn the terminology used in the field of study. Writing down the words often help in the thinking process in the sense it will extend your memory and you often have to write it in a way others should be able to understand. It is important to understand that the process of generating abstract thought from more basic terms is a time consuming process and it is not always easy to go the other way around. Like when you have seen a lot of examples of a phenomena, you can group it in an abstract term or sentence. This of course does not mean anyone reading this term or sentence will understand the same thing you do if they haven't seen the concrete examples. Which is why examples and concretization is so important.
   
6. The human mind is not a tool, but the thing performing the research using all of the above methods. Libraries are basically other minds thoughts collected, computer technology is a way to outsource procedures machines can do better than the human mind, and language the way we humans think. These tools can aid us, but not replace critical thinking, creativity ("wild" associations) and logical reasoning like deductive and inductive reasoning. Deductive is when you make conclusions from a set of assumed premises, and inductive is when you conclude based on a set of observations. You generalize. They could of course be mixed. Use inductive method to "verify" the premises.
   

What is good research

• Has a well defined problem (what we are going to figure out)
  
• Thorough enough (consider and care for as many variables as possible)
  
• Objective
  
• Eliminate unknown factors and adjust for known
  
• Gives new understanding of old knowledge
  
• Written to be easily followed
  
• Replicable by others
  
• Based on others work when applicable (Someone must have written something about this. Find it before you go ahead)
  
• Generalizable, point to a broader concept and is applicable
  
• Describes known limitations of research and sugest areas needing more research. (Creates new questions)
  
• Is possible within the economic and time limits you got
  
• Social acceptable (avoid harmful things, making deadly weapons, what consequences will it bring?)
  

What is bad research? (or not research at all)

• Repeating known information without adding something new (like new analysis)
  
• Biased: politically or commercially motivated (no hiding or removing of unwanted answers)
  
• Only look at the positive side when some of the science is both better and worse
  

Types of methods

• Descriptive research
  
• Historical research
  
• Correlational research
  
• Qualitative research
  
• Experimental research
  

Some principals:
We tend to conclude too early. Don't jump to a conclusion, reflect on it first. Try to figure out you bias.

Terminology:

Meta study Look at all available papers on a subject and try to make an average unified conclusion

Two assignments to think about till next time:
How would you go ahead to figure out a way (the variables and the method) to answer these questions?

Attitude campaign, does it reduce speed on a given road?
Reducing speed is a known way to reduce accidents, and what we actually want to know is if the campaign reduces the risk of accidents...

1. Would need a way to measure speed (average, number of violations), death count and accidents count
   
2. We need a control, like places with and without the campaign. We want multiple sites for comparison. They should be geographically separated, have similar death/accident/speeding statistics.
   
3. Then we compare after a long enough period. It is important to compare data at the same time of year.
   

Meditation makes better humans?
First you would have to define a “better human”, and since that question probably does not have an agreeable answer. There would also probably be way to many factors and causation would be extremely difficult to determine. Do people become better humans because of the meditation or are better humans more likely to meditate?

Lecture: Internet usage during lectures

The last part of the lecture was about a research paper "Internet devices as visual noise during lectures". The teacher had performed a questionnaire twice (2007 and 2011) to discover how internett connected devices used in class would function as attention grabbers.

• The paper does not give us an example of the questionare (although some of the questions are reflected in the paper)
  
• Does it say when the questionare was performed (during the semester)?
  
• In 2011: 39% had and used a computer. The remaining 61% were divided in about 1/3 (about 20%) not using a such device because the lecture did not facilitate it and the other 2/3 (40%) did not use it because they felt it reduced quality of lecture. (from graph)
  
• The text says that 57% of the none users gave "reduced quality of lecture" while the graph says about 65%
  

"It is therefore fair to say that when a majority of student still don’t use Internet connected devices during classes this is an active choice
they make. So, most students are somewhat concerned about the effects of using computers during class, and choose not to use one." That is actually not correct. About 40% had this opinion. The group saying they did not use an internet connected device because the lecture did not facilitate it might actually want to do so if changed. The subject would naturally be of great importance. A historical or discussion class might not facilitate the usage, but a programming or graphical design would

• Users and none-users reporting different values of academic vs none-academic use is natural. None-users would mostly focus on their notes or the teacher. Computer usage relevant or not but interesting easily catches the attention. When using a computer, more time is focused on their own screen. Student doing a lot of none-academic tasks might not want to admit it? Not only none-academic usage is relevant but also things that makes the screen change quickly and where in the classroom the device using users are sitting. Humans are biased in remembering bad or unwanted behavior.
  

The questions "entertainment-use of computers is very disturbing" and "Computer use does not affect me" 40% says computer usage does not affect them, and still 82% says none-academic usage is either somewhat disturbing or very disturbing. The second question should have sounded more like "Academic usage does not affect me"

The paper does not (and does not intend to) say anything about any changes in performance due to usage / none usage of these devices. It says nothing about the (possible) positive effects these devices can give.

---

[1] Practical Research - planning and design by Leedy and Ormrod