Study Guide for Test 2

Dr. Gass

HCOM 308

Fall 2008

Study Guide for Test 2

Chapter 8

perameter (wh= ole population), p.144

s= ample (subset of a population), p. 145

d= ata (records, observations, or measurements), p. 14= 6

d= escriptive statistics, (a.k.a. summary statistics), p. 146=

s= ingle numerical indicators that describe or summarize= data sets

m= easures of central tendency (averages), pp. 147-157

mean or arithmetic average, pp. 147-148

= most sensitive, precise measure of central tendency

= affected by extreme scores, e.g., outliers<= /span>

= may skew data

= most useful for statistical analysis=

= median, pp. 148-149

= physical middle point of a distribution, middle value = in a set of scores

= as many scores above as below the median

= often reported along with the mean

= less subject to skewing by extreme scores

= mode pp. 149-150

= used infrequently in social science research

= bimodal and multimodal distributions=

= unaffected by extreme scores, e.g., outliers

m= easures of variability (dispersion), pp. 157-165

= range, p. 158

= difference between highest and lowest scores

= variance (just the concept, not the formula), pp. 159-= 160

= standard deviation (just the concept, not the formula)= , pp. 160-162

= average amount of deviation in a distribution

= most sensitive, precise measure of variability

d= isplaying descriptive statistics in SPSS, pp. 151-154

Chapter 10

r= eliability defined, p. 188

Chronbach’s alpha, p. 192

Chronbach’s alpha in SPSS, pp. 193-194

acceptable ranges for Cronbach’s alpha, p. 195<= /o:p>

v= alidity defined, p. 202

a= measure can be reliable without being valid, p. 208

a= measure cannot be valid unless it is reliable, p. 209

Chapter 11

surveys vs. questionnaires vs. interviews vs. polls, pp. 213-214

&nb= sp;

d= o participants know the information? (p. 215)

= problem of “mindfulness”=

= problem of “non-attitudes”

= unreliability of self reports

w= ill participants disclose the information? p. 215

= social desirability bias, p. 208, p. 215

= intentional misreporting

p= opulation versus sample, p. 216

generalizability of results, p. 216

n= ominal level questions, p. 217

o= rdinal level questions, p. 218

i= nterval level questions, p. 218

r= atio level questions, p. 219

c= lose-ended versus open-ended questions, pp. 219-221

c= ross sectional versus longitudinal surveys, p. 221

p= ilot test, pp. 222-223

= 5-10% of sample

s= ample size

= rule of thumb 30-50 respondents per condition

= minimum of 5-10 subjects per condition

= larger sample size results in greater statistical powe= r

t= ypes and characteristics of interviews

= face-to-face, pp. 224-225

= telephone, p. 225

= self-administered, pp. 225-226

a= ssorted problems with survey research=

= response rate, p. 227

= unit nonresponse, p. 228

= item nonresponse (blank va= lue), p. 229

= nonresponse bias, p. 230

= improving response rates, pp. 230-231

q= uestions must be short, clear, succinct

a= void loaded questions (push polling)

a= void double-barreled questions

c= onduct a readability index (4^th-8^{th
grade level)}

a= void slang, jargon

s= ubtle variations in phrasing can affect perceived mea= ning significantly

a= void evaluative language

h= alo effects

r= esponse set

Chapter 13

p= retests and posttests, p. 259

a= ntecedent variables, p. 259

i= ntervening variables, p. 259

r= andom assignment, pp. 260-261, p. 268

e= qual likelihood principle

r= andom assignment matters more than random sampling

c= onfederates, p. 260, p. 263

w= ays of manipulating independent variables

= intact or self-selected groups

= stimulus conditions, p. 263

= stimulus videos, p. 263

= stimulus materials, p. 263

= hypothetical scenarios, role playing, p. 263

= problem solving, decision making

= manipulations must be substantial enough to produce a change, p. 265

f= actorial designs, pp. 264-265

= number of variables versus number of conditions

= 2 X 2 design, 2 X 3 X 2 design, etc., p. 264

= sample size

rule of thumb: 30-50 subjects per condition

minimum cell size 5-10 subjects, p. 265

= as the number of conditions increase, demand for subje= cts increases exponentially

m= easurement of the dependent variable

= scales and diagnostic instruments, p. 265

= behavioral measures, p. 265

= physiological measures, p. 265

c= ommon experimental designs

= Campbell and Stanley’s 3 types of experiments

= pre-experiments, pp. 273-274

= &nb= sp; one-shot case study

= &nb= sp; one group, pretest-posttest design

= &nb= sp; static group comparison &n= bsp;

quasi-experiments, pp. 274-276

&= nbsp; two group, pretest-posttest design

&= nbsp; time-series design

true experiments

= &= nbsp; pretest, posttest, control group design, p. 276

= &nb= sp; posttest only, control group design, p. 277=

= &nb= sp; Solomon four-group design, p. 278

Chapter 14

s= ample versus population

t= o be generalizable (p= . 283) a sample should be:

= random

representative

= sufficient in size

s= ampling frame: comprised of all the available participa= nts in the population

e= qual likelihood principle, p. 284<= /p>

p= robability sampling, randomly selecting participants from = the population, p. 284

central limits theorem, p. 284

simple random sample, p. 285

stratified sample, p. 286

cluster sdample, p. 287

systematic sample, p. 287

periodicity, hidden order to

nonprobability sampling , based on convenience, availability, p. 288

= convenience sample—most common type in social sc= ience research, p. 289

= volunteer sample, pp. 289-290

= purposive sample, p. 290

= quota sample, p. 290-291

= network sample, p. 291

r= eplication, pp. 293-294