Monday
Apr232012
Please Vote on the "Top Confusing Stats Terms"
Jeremy Taylor | Monday, April 23, 2012 at 7:01AM Please tell me what stats terms you think are the most confusing! Please order the terms you choose, according to how confusing they are (with #1 being most confusing). The results will dictate what topics are covered in future blogs!
Blog entries for Confusing Stats Terms #10, #9, and #8 are already posted, so I'm only asking for terms #7 through #1. Thanks for your input!
Related Content:
Top Ten Confusing Stats Terms Explained in “Plain English” (#8: Residual)
Top Ten Confusing Stats Terms Explained in “Plain English” (#9: Multicollinearity)
Top Ten Confusing Stats Terms Explained in “Plain English” (#10: Standard Deviation)
Editorial Note: Stats Make Me Cry is owned and operated by Jeremy J. Taylor. The site offers many free statistical resources (e.g. a blog, SPSS video tutorials, R video tutorials, and a discussion forum), as well as fee-based statistical consulting and dissertation consulting services to individuals from a variety of disciplines all over the world.
Reader Comments (3)
One of the most difficult aspects of teaching quantitative research to doctoral students is to first, foremost, and before anything else, understand who or what your population is. Before a sample can be taken, it is important to understand what borders guard the population, keeping errant individuals out and the important miscreants in. I would be interested in how others who teach statistics are getting the need to put borders around a population before sampling and subsequently hypothesis testing. Yes, I do understand that there are issues, problems, and more importantly detractors to hypothesis testing, but for goodness sake, get over it, quantitative dissertations would be even more boring without them, so help me understand how to explain to my doctoral candidates how best to "define" their population. All ideas welcome, good, bad, or statistically errant.
I am in need of a good explanation for eigenvalues so I would rate that as the hardest..a.t least for the time being. sorry I cannot think of anything else at this moment, but I'll let you know as they come to me. Love the concept of this blog. just stumbled upon it and I already know I will be here often especially because of my new analytics position I've obtained.
Elise,
I assume that you are asking about eigenvalues in the context of factor analysis, correct?
If that is the case, then an eigenvalue can generally be conceptualized as a measurement of the amount of variability in the data that a factor explains. As a general rule, an a factor with an eigenvalue of 1.0 or greater is considered useful (in exploratory factor analysis).
Thus, if you are deciding between a 3-factor model and 4-factor model and the 4th factor has an eigenvalue greater than 1.0, then you may lean towards the 4-factor solution (although the decision wouldn't be made on this alone). On the other hand, if the 4th factor has an eigenvalue of less than 1.0, then the usefulness of the additional factor may be suspect, so you may lean towards a 3-factor model.
I hope this helps!