Let's Talk About Stats, Baby...

Apologies for the late post. And the tone, I'm a bit punchy from the extra hours this week.

Ok, Lauer and Asher redeem themselves (probably because we're ascending the “experimental hierarchy”) with this chapter. They got me to do something I rarely do—read the appendix. And let me just say, I regret not reading that Statistics Demystified book I picked up last spring. They pack a ton of more or less understandable information about correlation into a small space. But enough about me being complimentary, time too whine, just a little.

The distinction between them [dependent and independent variables], however, is rather imprecise in descriptive research.” (86) Oh, please, get off it! Even from the examples they give it is easy to differentiate the characteristic that separates them: Independent variables are pre-existent and 'fixed,' whereas dependent ones are research subject-generated/derived.

Alternative Hypotheses. Nearly tautological explanation. (89) When using an alternative that is quite simply the inversion of the original hypothesis, how is this more valid than simply being honest about discovering negative results to the original one? Hmmmmm?

Oh, and yes, factor analysis kind of rocks.

Beach. Don't get me wrong, I love the Beach. But the only highlight that stood out to me was the point on p.237 about the limitations of content analysis being restricted to surface meaning because of the difficulty intuiting intention/motive of a writer. Since I am likely to be leaning in the direction of content analysis with my “online profiles” project, keeping this limitation in mind and finding methods to reduce its impact will benefit me (and perhaps a certain historian).