The flashcards below were created by user
zzto
on FreezingBlue Flashcards.


Experiment  vs. observation study 
 (imposes a treatment)
 (collect and analyze w/o change)
Can help determine cause and effect

Confounding factor
a variable in an experiment that was not anticipated before an experiment, but is known now (slope in fertilizer experiment)

correlation association
find definition

correlation does not mean
causation

Hypotheticodeductive reasoning
The Scientific Method


Experimental design:
(1) Replicate—(2) Assign treatments at random.(3) Statistical analysis is used to determine significant effects.

As the number of replicates increases, it becomes less likely that the
results were actually due to a variable that was not measured or controlled.

Assigning treatments at random helps to limit
the effects of unmeasured variables.

Three Basic Principles of Experimental Design
 The treatment is applied independently to the experimental unit(s)
 2 EUs)The treatment is randomly applied.
 The treatment is replicated…in space and time.

Stastics are
a way to quantify uncertainty

Statistics deals with
 Data collection
 Summarizing the data
 Placing data into some context

Statistics allows a scientist to make sense of  and to test 

a sample is a
subset of a much larger population

Two tailed test
two sided If the sample that is being tested falls into either of the critical areas, the alternative hypothesis will be accepted instead of the null hypothesis.(Ie boys may be smarter or girl may be)

Dispersion
(standard error, standard diviation, variance, rangemeasures values outside of mean)

Measures of central tendency examples
(mean median mode)

variance
the degree to which values deviate from the mean

The more deviation from the mean, the greater the degree of  in the data
“spread”

Alternative hypothesis (Ha)
the hypothesis that represents a change or an effect

Decision rule:
a rule for deciding whether or not to reject the null hypothesis

Null hypothesis (Ho):
the hypothesis that represents no change or no effect


Error
When our hypothesis is wrong

Type I error is erroneously saying things are
different when in fact they are not.

S^{2} _{p}
is the pooled variance; assumes both populations have equal variances



In order to compare the 2 samples (female vs. male) we need to
 Compute the t statistic based on our data.
 Determine the degrees of freedom.
 Set the level of significance (α).
 Compare tcalculated statistic to tcritical statistic.
 If t_{calculated} > t_{critical}, REJECT H0

