Experiment 2, Lab 2

Tasks:
1. Inspect the class dataset in SPSS
2. Run t-tests on the results
3. Make bar graphs and tables of the results – and save them
4. Discuss what the results tell us

1. How to Open SPSS and Inspect Your Data

We will be using SPSS to examine the mean results from the entire PSYCH100 class.
Download the dataset from Moodle (an SPSS file). Your Lab Instructor will show you where.
Notice that below the file name on Moodle there are some summary statistics about the
sample of people who completed the experiment – you will need that information when it is
time to write the Method section of your lab report.

Double click on the file to open it. SPSS should open automatically. If it doesn’t, open the
Windows Start Menu (bottom left of the screen) and select SPSS which is listed under P for
Psychology. If you do not see SPSS there, use the search function to find it. From there you
can find and open the data file. It will probably look something like the screenshot in Figure
2 (but with lots more data). Each line/row in the SPSS file is a different person’s mean
results.

Figure 2: Example class data file in SPSS

2. Run a t-test on the results

What we want to know is whether the different types of words make a difference in the
speed and accuracy of the colour naming. Although the nonsense word condition provides
an interesting control condition, for the current analysis we are simply going to compare the
colour name words and common words.
In Experiment 1 we determined whether there was a numerical difference between our
conditions. But in Experiment 2 we want to know whether our two conditions are
statistically different (or whether the differences could have been due to chance). To know
this, we need to run inferential statistics. All of the participants saw both levels of the
independent variable (word type) so we are going to use a Paired-Samples t-test (if we had
shown each word type to different groups of participants, it would be called an Independent
Samples t-test). We are going to run two t-tests: one comparing reaction time for the Colour
name and Common words conditions, and the other comparing accuracy for those
conditions. We will set up both t-tests at the same time.

To run the t-tests, open Analyze from the top menu bar of SPSS, then choose Compare
Means, then Paired-Samples T Test (see Figure 3).

Figure 3. Paired Samples t-test selection

In the box that opens, move the two response time variables (Common words, Colour
names) to the Paired Variables box on the right using the arrow (see Figure 4). Now do the
same with the two accuracy variables (Common words, Colour names), and click OK and
take a look at the output.
 Figure 4. Defining our pairs

The first table in the output labelled Paired Samples Statistics provides descriptive statistics,
which will show you the mean and standard deviation.

Scroll down to the table labelled Paired Samples Test, which gives us the outcome of our
paired samples t-test. It will look similar to the one in Figure 5, though of course your
numbers will be different. There is one row for each test that we ran (one comparing
reaction time, and one comparing accuracy). In this table, in the first column you will see the
difference in the mean between our two comparison conditions.

Figure 5. Output of the paired samples t-test

The three columns at the end of this table provide the results of our t-test.
● t is our test statistic – the further away from 0 this number is, the more likely it is
that our conditions are statistically different (this number is used by SPSS to calculate
the p-value).
● df is our degrees of freedom, which for a t-test will be our sample size minus one.

● Sig is our p-value, which gives us the probability that our results are due to chance.
SPSS will give you p-values for one-tailed and two-tailed tests. For this comparison
we want the p-values for the two-tailed tests.
● If the p-value for a given test is smaller than our threshold (in psychology, generally
0.05), then we would say that there is a statistically significant difference between
 our two conditions—meaning that the observed pattern of results is unlikely to have
occurred by chance alone. But if our p-value is larger than our threshold (0.05), we
would say that the two conditions are not significantly different from each other—
meaning it is unlikely that there is any real difference between our two conditions. In
Figure 5, our p-value is so small that SPSS is not giving us an exact number, and
instead reports it as less than 0.001, indicating that our conditions are significantly
different from one another.

We should include these three numbers when we are reporting our results: the test statistic
(t), degrees of freedom (df), and p-value (p). We report these in the format: t(df) = X, p = X.
You want to report your results in words first, then provide the statistics. For example, we
would report the results for reaction time from Figure 5 as: “The mean reaction time for
Common words was significantly faster than the mean reaction time for Colour name words
(t(277) = -14.998, p < .001).”

Make a bar graph and table of the results

Bar graph
Just as we did in Experiment 1, open Graphs in the top menu bar. Select Legacy Dialogs and
then Bar. We want a Simple bar chart, then tick Summaries of separate variables as shown
in Figure 6.

Figure 6. Bar chart options in SPSS

Click Define. Select both of the reaction time task types (Common words and Colour names)
and move them from the list on the left to the right Bars Represent field using the arrow.
Click on Options, then tick Display error bars. The type of error bar should be confidence
intervals. Once you have all the variables selected and options chosen, click on Continue
then OK.
 Figure 7. Graph produced in SPSS

You should see a graph that looks something like the graph in Figure 7. As with Experiment
1, the bar graph produced by SPSS requires editing to meet the APA requirements for
figures. A reminder, the main things to do are add more informative axis labels on both
axes, remove the gridlines, and remove the colour fill on the bars. You can find detailed
information about the formatting requirements for figures on Moodle. We will be more
strict on APA formatting for Lab Report 2 than we were for Lab Report 1.

To save your graph for putting into your lab report later, select the graph, then right click,
then Copy As, then Image. Open a document (e.g. in Word) and Paste into the document as
an image. Then save the document with an informative name (like “PSYCH100 experiment 2
graph”) and either email a copy of the document to yourself or save it to your University
WebDrive/WebHome (H: Drive) – you can read more about accessing your H: Drive in lab
handout 1.2.

Now repeat the steps above, this time looking at accuracy instead of reaction time.

Table
Next we will calculate the mean reaction times for each type of task, so you can put them in
a table in your report. It is tidier to put both reaction time and accuracy variables in the
same table. Open Analyze on the top menu bar. Select Descriptive Statistics and then
Descriptives. Move your reaction time and accuracy variables (Common words, Colour
names) from the left field into the right field using the arrow, and then click OK. You should
now see a descriptive statistics table for your four variables, see Figure 8 for an example.
 Figure 8. Descriptive Statistics table produced in SPSS

Like the graph, this table requires editing to suit APA formatting. Again, you can find more
detailed information about APA formatting requirements for tables on Moodle. Rather than
editing the table in SPSS, it is easiest to copy the data from the SPSS table and paste it into
Word, Excel, or similar program with table-editing functions, where you can then format the
table as required. A reminder that the main things to change are removing the rows and
columns you don’t need (you only need a row for each task, and a column each for the
mean and standard deviation), removing all vertical lines, coming up with more informative
column headers, rounding to 2 decimal places, and removing all colour formatting.

You may run out of time to finish working on your table during the lab – that is ok. As long
as you have saved the data, you can finish formatting it in your own time. To save your table
for putting into your lab report later, save the Word or Excel document you have been
working in with an informative name (like “PSYCH100 experiment 2 table”) and either email
a copy of the document to yourself or save it to your University WebDrive/WebHome (H:
Drive).

4. Discussing your results

Your Lab Instructor will lead you in a brief discussion about today’s tasks.
● Do our results match your hypotheses from last week? Explain how/why.

● What other examples of automatic cognitive processes can you think of?

● What tie-ins with the lecture content can you see?

Before you leave: Make sure you have completed the SPSS analyses and saved copies of
your final graphs, the data to go in your tables, and the results of your t-tests, somewhere
you will be able to find them again later. You should also now understand what we found,
and what that pattern of data means.