Biggest problem and learning more
Вставка
- Опубліковано 24 сер 2022
- I was recently asked two interesting questions. During the 2022 AoM Meeting, I was one of the experts in meet the quantitative expert session. The idea of the session was to give the participants a chance to ask questions. Normally the questions are about problems that the researchers face themselves, but this time the first question was: "What is the biggest problem in methods in management research?"
The second question came in in an email yesterday and prompted the recording of this vlog. A student from my advanced research methods course asked: "How do I learn more about methods?". My standard answer to the question is to come to my basic course (one semester), and then my advanced course (one academic year) and read the course materials. What made this question unique is the context: it came from a student who had already taken all the training that I offer. So in the second half of the video I talk about how to train yourself beyond what the university has to offer.
Note: The explanation of two-stage structural equation modeling (TSSEM) is incorrect in the video. The technique does form a large meta-analytical correlation matrix contrary to what I claim.
Really great advice! I've watched most of your videos, and they have really functioned as a backbone for learning methods for my PhD. I would also add one piece of advice. Learn to programme the models from scratch using Python and online tutorials. This really forces you to understand what the model does, and it functions as a great exercise to learn programming, which then makes it easier to explore new methods, such as sentiment analysis, and calculate more complicated variables, such as those used in social network analysis.
Right. I do not use Python, but I make students to implement things themselves. We start with OLS regression and how to minimise the sum of squares using excel. Here is a screencast from years ago: drive.google.com/file/d/0BwkyLuHVQtWbeGgxUkRIT0J5eW8/view?usp=sharing&resourcekey=0-HOqaCOi7i4PvquZnhOhHAw
@@mronkko hello professor, the youtube link you have provided for the screencast is the link for this video instead. Would you mind share the correct link again? Thanks
@@francischuahcw What screencast are you referring to?
@@mronkko hello professor, in your above response, you stated that "We start with OLS regression and how to minimise the sum of squares using excel. Here is a screencast from years ago: ua-cam.com/video/m5qiZgsmNQA/v-deo.html"
The youtube link which you have provided, redirect me to this video again. I supposed the link should direct me to the video where you talks about minimising the sum squares using excel?
@@francischuahcw Got you. This is the correct link: drive.google.com/file/d/0BwkyLuHVQtWbeGgxUkRIT0J5eW8/view?usp=sharing&resourcekey=0-HOqaCOi7i4PvquZnhOhHAw
I plan to redo the screencasts for the next run of my course and put them to the channel.
Thanks for sharing your knowledge. Your videos has been really useful for my academic career.
You are welcome
Thanks Mikko. Just a question: what was the name of the IV method based on heretoscedasticity? And could you share a reference for the method?
Ebbes, P., Wedel, M., & Böckenholt, U. (2009). Frugal IV alternatives to identify the parameter for an endogenous regressor. Journal of Applied Econometrics, 24(3), 446-468. doi.org/10.1002/jae.1058
I do not know if these techniques have specific names, but if you put
instrumental variable heteroskedasticity
into Google Scholar, you will find lots of links. I have a video about these on my list of things to do.
Thanks Mikko, I am looking forward to seeing your video. This technique sounds extremely exciting.