NEUR 2001 Study Guide - Midterm Guide: Blind Experiment, Signal-To-Noise Ratio, Confounding

118 views4 pages

Document Summary

Research sampling: general terms, population: entire set of individuals that share a common characteristic of interest, sample: subset of population, used to make inferences/generalizations about pop. *representative sample resembles pop: biased sample, def: diff characteristics than pop of interest, occurs via chance or selection bias. Selection bias: def: selection method that inc the probability of selecting a biased sample, causes: survey return rate, door-to-door, highly motivated participants, $ compensation. *ppl (cid:449)ho take the ti(cid:373)e to fill out a su(cid:396)(cid:448)e(cid:455) a(cid:396)e auto(cid:373)ati(cid:272)all(cid:455) diff tha(cid:374) those (cid:449)ho do(cid:374)(cid:859)t. Goals: optimal size: ensures stat significance w/o wasting resources & animals, larger = more representative. Esou(cid:396)(cid:272)e e(cid:395)(cid:859)(cid:374: e = total # animals total # groups, goal: 10