Evaluating Recorded Sounds

As I am currently measuring the study environment’s possible effects to the study results, I have to find a way to evaluate the latter. The outcome of my online course’s 1st level weresupposed to be four recorded exercises:

  • Ex. on open strings (video example only)
  • Scale from G to G on solo strings (graphic, notation, video)
  • A folktune from standard notation (notation, audio)
  • A simple melody by ear (audio example only)

I plan to give 1 to 5 marks to every exercise. The first question I have right now is: should I really take all four into account or would the scale be enough? I guess if I take all four, it cannot make anything worse but it is just much more work. Also, since the exercises were presented in different ways, I could get some interesting connections between some environmental elements and certain exercises.

The second question is about evaluating: which characteristics should I evaluate? In this stage, every student is struggling to play the correct notes with the given rhythm so there is really no point to evaluate any musical aspects. So these characteristics would be:

  • Correct notes
  • Correct rhythms
  • Stable tempo
  • Evenness of tone (volume, sound)?

And the scale would be a 5 step Likert.

Measuring Frequency on Likert Scale

I’m putting together a questionnaire for my survey. I was thinking what is the best way to let the respondents to describe their study environment. And I came up with asking the “how often?” question.

For example:

How often…

  • … were you alone while studying?
  • … did you study in silence?
  • … did you study in front of the computer?

etc.

Here’s why I ask the frequency and do not let people to describe their surroundings:

1. From the former studies I know that it took several study sessions to complete the first level of the course. So if I would ask How crowded was the room? or How many people were in the room? then the respondent would not know which session to describe. Some people would naturally describe an average situation but not everybody.

2. A factor can either distract you from learning or not (in the CLT point of view – something is either using your cognitive resources or not, taking your attention or not). Measuring the amount of distraction would be too difficult and more to the qualitative side. So there is no point asking “How much”. Instead I ask “How often” to get the amount of distractions.

3. I know that measuring frequency on the Likert scale must be done very carefully. But right now I cannot imagine having numeric choices – once a week, once in a lifetime, every other second. People took the course in different tempo. The number of sessions was different and the time span too. After all, some people did it a while ago so very specific questions would be rather annoying and the answers not reliable. Besides, I have “never” at one side of my scale and “always” on the other side. Only if everybody answer “never” to all questions can I argue that the environment had now effect to their learning.

So here’s my Likert scale (in Estonian):

  1. mitte kunagi (never)
  2. harva (rarely)
  3. mõnikord (sometimes, occasionally)
  4. sageli (frequently)
  5. alati (always)

I browsed through many examples but the language kicks in: the meaning of sometimes in Estonian (mõnikord) is rather less than a half. But I want to have an equal scale with a proper center. Still, since I will have to translate it back later, I have chosen the “mõnikord” (sometimes).

To have a numeric scale that is even by default? No, that would be describing time through other numeric scale that we are used to (seconds, minutes, hours etc).

 

 

Effects of the Environment When Starting to Play a Musical Instrument Instructed by a Multimedia Study Material

Studying a musical instrument with the help of a multimedia study material is different from studying any theoretical discipline because there is an extra source of information: the additional visual and audio information that the learner creates by playing the instrument.

As the learner’s cognitive resources are limited (CLT), there is a higer risk of cognitive overload. But before starting to optimize the study material we should see what else can use learner’s cognitive resources.

250+ beginners studied guitar with the help of the www.guitarschool.ee multimedia course and recorded 4 test exercises. The study process took place in a natural environment (75% home, 20% office, 5% elsewhere) and was not controlled (most common case when studying a musical instrument). The study material was the same for everyone so it is possible to measure the effect of the study environment on study results. If the effect is significant then it is important to reduce it in order to improve study results. Otherwise whatever improvements we make to the study material to use of the learner’s cognitive recourses more efficiently, the overall effect may be not big enough.

It is not possible to change the actual environment, we only can change the study material. There are two ways to neutralize the effect of the study environment by the study material:

a) reducing the cognitive load caused by the study material. This means putting through less information and causing the study process to slow down.
b) reducing the effect of an environment by demanding more engagement from the learner (gamification, more interaction)

I will need 3 bits of information:

1) Facts about the environment (What was it like? from questionnaire)
2) Cognitive effects of the environment (How did it feel? from questionnaire)
3) Study results (recordings, achieved level of total 8 levels of the course)

The first set of questions asks the respondents to describe the actual characteristics of an environment, the second set wants to know how did it feel. For example, if there were potentially annoying factors in the room but the learner didn’t notice them, then we cannot consider them as extraneous load (CLT).

Planning a Case Study

I started my online guitar course at www.kitarrikool.ee in May 2007. Now,  5 years later it has 11 000+ registered users. Of course, not all of them are studying. The easiest way to determine whether someone has actually learned is to see if they have uploaded their recorded exercises.

Sending the recordings of the course’s first level has always been free. The rest of the online course w feedback is free for the book owners. There are many users who learned but didn’t send exercises.

Facts

  • Total amount of registered users – 11077
  • Uploaded files – 2502
  • Sold books – 1900 (192 registered their book to enable the feedback feature)
  • Users who uploaded anything – 430
  • Users who studied with the help of my materials and did not attend my classes before uploading at least the full 1st level and (4 exercises) – 233 (75 female, 157 male)
  • The course is made up of 8 levels
  • Number of control exercises per level to be studied and recorded – 0 (8th level) to 4 (1st level), mean number of control exercises being 2.625 per level
  • Number of feedback e-mails written – 1000+

Why e-learning and how is it designed to work? If questions are WHY and HOW, then a case study can help (Yin, R. 2003)

The higlighted 233 people are the ones who can help to get a picture of how and why an online musical instrument course works. The next question is: what to ask from those people?