As promised (or perhaps forewarned), today’s focus was on quantitative analysis and taken by Dr Mira Peter. She faced teaching in a. Computer suite that had high ceilings and bare walls – it reminded me of my teaching space! Regardless of the environmental challenges, I learned a great deal that I had found challenging when I read it. I won’t type up all seven pages of notes (just wanted to throw that out there), but here are the key ideas I took away.
- There are four common quantitative data gathering methods – observer action, interviews, questionnaires and databases.
- A sample (which should be representative of the wider population in order for generalisation to occur) which is not representative is called BIASED
Some aspects were a throwback to high school statistics (thank goodness I had that foundation), and other aspects were new. Key terms included
- Variable (something that changes)
- Statistic (collected from the data
- Parameter (statistics when talked about in terms of the population as a whole – inferred population)
There are two main types of variables: Independent (which is manipulated by the researcher) and dependent variable (DV) which is the OBSERVED OUTCOME (data of the independent variable being manipulated. Extraneous variables are also important, consisting of those variables which may affect the outcome e.g. Age of participants, experiment, the researcher or environmental factors. Extraneous variables need to be controlled for as much as is possible.
There are three kinds of data (at least – these were the three that were discussed in terms of today’s session): Nominal (yes/no type answers), ordinal (choose from one to five on a scale type answers) and continuous/scale data (think timing an athlete, measuring height etc).
We also learned how to work out a standard deviation (and then how to do it on excel), and about confidence intervals. Effect sizes were also discussed and how to work these out. And the different numbers – what they mean.
Alright… a better sentence about that….
There’s this thing called Cohen’s standard which is used by statisticians and data analysers. If (after various calculations are made the effect size is…
- 0.2 to 0.4 the effect size is small – real but hard to detect.
- 0.5 to 0.7 the effect size is medium – it can be seen and noticed
- 0.8 or above the effect size is large and very very obvious.
If the effect size is one, in education, this means that more than one year’s growth has been achieved.