Understanding Psychology (10th Ed)

R E C A P / E V A L U A T E / R E T H I N K
K E Y T E R M S 
R E C A P 
How do we generalize from data?
• Inferential statistics—techniques that use data from sam-
ples to make predictions about a larger population—are 
useful in deciding whether differences between distribu-
tions of data are attributable to real differences or to 
chance variation. (p. A-14) 
How can we determine the nature of a relationship, and the 
signifi cance of differences, between two sets of scores?
• Measures of relationship provide a numerical index of 
the extent to which two variables are related. The corre-
lation coeffi cient ranges in value from 
1
1.00 to 
2
1.00, 
with 
1
1.00 indicating a perfect positive relationship 
and 
2
1.00 indicating a perfect negative relationship. 
Correlations close to or at zero indicate there is little or 
no relationship between two variables. (p. A-16)
E VA LU AT E
1.
Researchers would like to estimate the level of stress for 
fi rst-year college students for a given year at a large uni-
versity. A stress index is given to a randomly assigned 
group of 500 fi rst-year students. The class size is 6,000 for 
that year. In this example the group of 500 is known as a 
, and the entire class of fi rst-year students is 
known as the 
.
2.
Dr. Sanders states that the results of his experiment show 
a difference between the two groups, and that there is a 
90% probability that the results are due to a true difference 
between the groups and not due to chance. Are his results 
statistically signifi cant in the way the term signifi cant typi-
cally is used by psychologists?
population p. A-14
sample p. A-14
inferential statistics p. A-14