Time flies and it's down to the second last session. Dr. Quek started the lesson by reviewing what we have learnt in the module so far, which also served as the context for the topic that she was going to talk about.
She showed us a research involving the impact of the use of Classroom Performance Systems in the class done by a beginning researcher and based on the information given, I found it difficult to see the alignment between the rationale, objectives and the research questions of the study. Dr. Quek clarified that the researcher was an amateur and that he had since amended his study. She added that she will pass us a copy of a paper presented by Dr. Sue Trinidad. Furthermore, when we embark on any study, we must first craft the research questions first (which usually takes months), and we must know we can measure the outcomes.
Next, she highlighted that when we are modifying an instrument, we can actually just pick up the relevant scales and not use all the scales in it. Furthermore, we can also combine scales found in different instruments. One example that she showed us was a study done using scales from the WIHIC as well as a scale called
material support, which was not from WIHIC. However, she cautioned us that the instruments used must be validated, reliable, adhere to Moos' theoretical framework before we can combine the scales from them. Furthermore, we need to be careful not to combine QTI with the other instruments we have learnt so far, as QTI measures teacher-student interactions.
Karen clarified with Dr. Quek about how we will know how well an instrument measures something, to which Dr. Quek said that the alpha reliability must generally be greater or equals to 0.70. However, she added that if the alpha reliability goes beyond 0.95, then it may be "too good to be true". Piqued by the alpha reliability, I did a Google search and found two websites which seemed to explain it quite well. They are:
http://www.ats.ucla.edu/stat/Spss/faq/alpha.htmlhttp://www.jalt.org/test/bro_13.htmFrom the first website, I also found that we can compute the alpha reliability using the following formula

However, even though an example was given, the explicit calculations to obtain the alpha relaibility were not shown.
As promised, Dr. Quek passed us each a copy of Dr. Sue Trinidad's paper. We were asked to:
- read and identify the gaps found in the dimensions of the classroom
- discuss the improvements made
- evaluate if the gaps and improvements were effectively addressed
However, there were some alignment issues with the printed copy so I went online to find the paper and uploaded the correct version to the Lesson Box. The file was downloaded from
http://www.aare.edu.au/05pap/tri05058.pdf.
Even though the write-up was relatively long, it was surprising readable. It also helped that I had three very enthusiastic group mates to chew over and digest the information. However, Dr. Trinidad seemed to use the term
Computers in Education and
Computers in Teaching interchangeably, which was a huge source of confusion for us. In addition, we had difficulty trying to pick up the information required. When we were done with our group report, Dr. Quek introduced another useful tool, Google Docs, to us. We were asked to create a powerpoint presentation and share it with the rest of the class. In no time, I have gained access to all the other three groups' work! Everything is stored online and I can even see who has modified which document. More importantly, it's free! Power!
We were next introduced to the paired-sample t-test. Even though I had learnt this way back in JC as well as in MED871 and MED872, I struggled when my classmates asked me about it. It seems that most of my classmates did not know this test at all. I guess the greatest test of whether you know something or not is to see if you are able to explain it convincingly to people who do not know it at all. I did a search and found the following:
Statistics GlossaryAn example of how the test is usedDegrees of freedomI guess I really have to re-familiarize myself with my statistics if I want to embark on any research then.
After lunch, we were given WIHIC (both the preferred and actual forms) to do. We then compiled our results in a template given by Dr. Quek, first in our group and then as a class. As the template had provisions for more items in a given scale then were present in our survey forms, we deleted the additional columns, not realising that the formula would be affected. Yuen Han pointed that out later on and as a result, my group had to use the results compiled by them instead. Important lesson learnt!
After we have compiled our results using EXCEL, we copied them into SPSS. The original version installed on our desktops was 15.0, but we realised that the license has already expired. All our hard work went down the drain as we could not even save anything. Luckily, a helpful technician was on-hand to help us install version 16.0 and before long, we were up and running. I think this incident really reinforced Dr. Quek's point about having lessons in NIE instead of Teachers' Network, as resources and infrastructure are much better over at the former, even though Easterner me still thinks NIE is too far!
Anyway, we realised that there were some bugs with SPSS and we had to manually overwrite them. We then churned out the results and exported them to html form, so that we can view them from any web browser without the need for SPSS. My group churned out a total of 12 files (one for each scale, actual preferred, and one for the paired-sample t-test). When I saw what Yuen Han's group has done, I realised that you can actually combine the reports into one neat little html file! Another lesson learnt! Should always play around with the different functions.
After we were done exporting the results, Dr. Quek gave us some homework to do. We needed to extract the results from the reports and compile them into a table. Always nuturing and kind, she added that we can always compare our tables in the next lesson to see if we have done anything wrong. I guess that was really reassuring for me as well as for many of my classmates, who also struggled with SPSS.
P.S: SPSS stands for Statistical Package for the Social Sciences, and the current version is 17.0!Before we ended the lesson, Dr. Quek reminded us that Thursday will be our last session and we will need to submit a report on that day too. It really seems like yesterday when I barged into the class not understanding a single thing. I really must thank Dr. Quek and my wonderful classmates for helping me catch up. Nice people do exist! I hope all of us will do well for the module. =)