Page 21 - 2023-ml-lewandowski-stats1e

For participants in the control condition, we have them watch a sim-
ilar-length video featuring brown noise. To measure our dependent
variable, we have all participants complete the Perceived Stress Scale
(Cohen et al., 1983), which asks participants to respond to items
such as, “In the past month, how often have you felt nervous and
‘stressed’?” Participants respond to each item on a 5-point scale, with
0 indicating that they never feel a certain way and 4 indicating that
they feel that way very often. (If you want to test your own stress level,
enter the search terms Perceived Stress Scale to access the measure.)
Now that we have our study design, imagine that we recruit 80
participants, half of which we randomly assign to the mindfulness
condition and half of which we randomly assign to the control condi-
tion. After completing the intervention phase of the study, we measure
our two groups’ stress level and find that the mindfulness condition’s
mean is 2.56 (SD = 1.08), while the control condition’s mean is 3.148
(SD = 0.93), with higher scores indicating that participants are expe-
riencing higher levels of stress. From this information, does mindful-
ness meditation reduce stress? Well, we don’t know yet. Although the
mindfulness mean is lower, it could be that the differences between
the two groups’ means are due to sampling error. In order to say that
there is an effect of mindfulness meditation, we need to conduct the
t-test for two independent samples.

t-Test for Two Independent Samples: What We’re
Trying to Accomplish
Our research goal is to determine if participants who engage
in mindfulness meditation represent a different population
from, or the same population as, people who don’t engage in
mindfulness meditation. If the mindfulness sample represents a differ-
ent population, they should report significantly lower stress levels than
participants in the control condition. To test this, we will build a com-
parison distribution starting with the hypothesis that mindfulness med-
itation has no effect. We will then evaluate just how probable it would
be for us to observe a difference between the means, if mindfulness has
no effect (this is what the p-value tells us). Thus, the key question we
will evaluate is whether the difference between means that we observe
in our study is expected or unexpected, assuming the null is true.
Hypothesis Testing with a t-Test for Two
Independent Samples: How Does It Work?
To conduct our hypothesis test, we need to see where our
study results fall on the distribution of differences between
means. Before we get to this step, though, we need to start
by making sure we meet the assumptions of the test.

T -TES T F OR TW O INDEPENDENT/UNRELATED S AMPLES • CHAPTER 10 363

Copyright ©2024 Worth Publishers. Distributed by Worth Publishers. Not for redistribution.

11_statsresandlife1e_24717_ch10_343_389.indd 363 29/06/23 5:17 PM