Creative Commons license icon

Reply to comment

*laughs* Sorry about that; for the most part, those can be ignored. That's my attempt to "show my work", and represents a compromise between furs who say "just give me the findings in plain language" and those who say "you can't just make claims like that... show us your statistics!" I feared that starting off the article explaining the rationale of t-tests and regression analyses would intimidate furries who just peeked in to read a short summary of what we learned (and who want to read something in plain English).

Greenreaper did a pretty good job explaining it =) Basically, if the p-value is "less than .05", it suggests that:

a) if it's a t-test, a difference between means of that magnitude would only be expected to be due to random error less than 5% of the time.

b) if it's a regression analyses, a linear relationship of that magnitude between two variables would only be expected to occur less than 5% of the time.

All of this needs to be taken with a grain of salt, of course. Statistics are based on assumptions and are inferences about reality based on a limited sample (Staghorne does a good job describing some of these concerns). That said, this IS how the vast majority of social science (and "hard science", for that matter) is analyzed!

Reply

  • Web page addresses and e-mail addresses turn into links automatically.
  • Allowed HTML tags: <a> <img> <b> <i> <s> <blockquote> <ul> <ol> <li> <table> <tr> <td> <th> <sub> <sup> <object> <embed> <h1> <h2> <h3> <h4> <h5> <h6> <dl> <dt> <dd> <param> <center> <strong> <q> <cite> <code> <em>
  • Lines and paragraphs break automatically.

More information about formatting options

CAPTCHA
This test is to prevent automated spam submissions.
Leave empty.