“Enough with the lecturing” at
This NSF study “proves” that active learning is better than just lecturing.
The devil is in the details.
The quote from the sited “study”: “Some 225 of those studies met the standards to be included in the analysis including: assurances the groups of students being compared were equally qualified and able, instructors or groups of instructors were the same, and exams given to measure performance were either exactly alike or used questions pulled from the same pool of questions each time.”
And another quote:
“Freeman and his co-authors based their findings on 225 studies of undergraduate education across all of the "STEM" areas: science, technology, engineering and mathematics.”
To meet the requirements of an actual scientific study all “sub-studies” must had used THE same (or convertible) measuring instruments and procedures.
Clearly, measuring tools and procedures depend on the teaching subject. However, that is not an actual issue.
The original paper (http://www.pnas.org/content/111/23/8410.full) basically says “we took 255 published papers and ran various data analysis, and this is our conclusions”.
There is no way to follow the logic which lead from data to conclusions (for example, in math sciences, if one proves a theorem, everyone else can trace all the steps, just in case – for a human factor – if a mistake was made – that IS science).
The most important issue with this type of meta-analysis is that it based on the data which initially do not fulfill the requirements of a scientific study.
We read that “those studies met the standards to be included in the analysis including: assurances the groups of students being compared were equally qualified and able, instructors or groups of instructors were the same, and exams given to measure performance were either exactly alike or used questions pulled from the same pool of questions each time.”
This statement is very misleading.
When we read it, it sounds like all study in physics were based on using the same measuring instrument and procures; or all studies in each another subject area also were based on using the same measuring instrument and procures.
If that was a case that would be grate!
In reality, it only means that authors of one paper on effects of active learning in physics used the same (or similar) exams when testing studies in a controlled group and in the experimental group. But other researchers in a different study were using exams (or other assessing techniques) different and incomparable with ones used in other studies.
However, if we examine each individual study we find that they do not sustain the same scrutiny we would apply to a study in physics, mathematics, or any other STEM field.
An example of such analysis is available here: “Critical Reading of “Making Sense of Confusion” by Jason E. Dowd, Ives Araujo, and Eric Mazur” at http://gomars.xyz/msm.html or http://www.scipublish.com/journals/EPI/papers/1404.
If all individual studies do not satisfy criteria for a scientific study, and if for all individual studies the results can be explained by other reasons than active learning, the conclusions of the meta-analysis have no solid basis.
It does not mean that I state that active learning does not work. I only say that we have no solid evidence for making the statement that active learning is better than a lecture.
And there are also much simpler explanations for an effect provided by active forms of teaching, for example, “the first law of TeachOlogy”: “If we take two large groups of similar students, and one group of students will have a more extensive or diverse learning experience (for example, more contact hours, or more time spent on certain exercises, or training through more and different exercises, etc.) students from that group, on average, will demonstrate better learning outcomes than student in the controlled group.” (more at http://www.teachology.xyz/WNSF.html and http://www.teachology.xyz/6LT.html).