Professional Documents
Culture Documents
09
Modul ke:
Uji Hipotesis
Fakultas
Psikologi Niah Maretno Sari, M.Psi.,Psikolog
Program Studi
Psikologi
THE ROLE OF INFERENTIAL STATISTICS IN RESEARCH
• a random sample may be more or less representative of a population because, just by the luck of the draw,
the sample may contain too many high scores or too many low scores relative to the population. Because
the sample is not perfectly representative, it reflects sampling error, and so the sample mean does not
equal the population mean.
• Researchers perform inferential statistics in every study, because it is always possible that we are being
misled by sampling error so that the relationship we see in our sample data is not the relationship found in
nature
• Previously we’ve said that inferential statistics are used to decide if sample data represent a particular
relationship in the population.
SETTING UP INFERENTIAL PROCEDURES
• The mean of the sampling distribution always equals the of the µ raw
score population that says we are representing.
• A sampling distribution always describes the situation when Ho is true.
Here it shows the sample means that occur when we are drawing samples
from the IQ population where µ is 100. Any sample mean not equal to 100
occurs solely because of sampling error—the luck of the draw that
produced an unrepresentative sample.
Determine the size and location region of rejection
• Choose alpha : Recall that the criterion probability defines sample means as
being too unlikely to represent the underlying raw score population, which in
turn defines the size of the region of rejection. The symbol for the criterion
probability is α, the Greek letter alpha. Usually the criterion is 0.05 , so in
code, a α = 0.05.
• Locate the region of rejection : Recall that the region of rejection may be in
both tails or only one tail of the sampling distribution. To decide, consider
your hypotheses. With our pill, we created two-tailed hypotheses for a two-
tailed test because we predicted the pill might raise or lower IQ scores. We
will be correct if our sample mean is either above 100 or below 100 and we
can reject that it represents the no-pill population.
• Determine the critical value: We’ll abbreviate the critical value of z as z crit .
With α = 0.05 , the total region of rejection is 0.05 , of the curve, so the
region in each tail 0.025 is of the curve. From the z-table, a z-score of 1.96
Computing Z
Interpreting significant result
• Significant does not mean important or impressive. Significant indicates
that our results are unlikely to occur if the predicted relationship does not
exist in the population. Therefore, we imply that the relationship found in
the experiment is “believable,” representing a “real” relationship found in
nature, and that it was not produced by sampling error from the situation
in which the relationship does not exist.
THE ONE-TAILED TEST