Instructions for lab report

For this lab report, you will analyze the group data from the lab exercise, plot the group data means (provided by the instructor) on the graphs in the MEL Lab workbook, and write a lab report in APA format, emphasizing the method and results sections. The graphs from the workbook should be included with your lab report as figures.

From the Data Files link on the syllabus you will find two files for Lab 2.10, lab2_10_data.txt and lab2_10-tables.txt.

The file lab2_10-tables.txt contains the means for the two tests (recognition and recall), first averaged across all participants, and then for each participant individually. Use these means to create the figures (line graphs) in your workbook. You will include those graphs as figures in your lab report.

The file lab2_10_data.txt contains the data for analysis. Each participant’s data is on one row, and there are 5 columns. Column 1 is the subject number. Column 2 is the proportion correct in the recognition test (c_recog). Column 3 is the proportion of intrusions in the recognition test (i_recog). Column 4 is the proportion correct in the recall test (c_recall). Column 5 is the proportion of intrusions in the recall test (i_recall). The first row contains variable names for each column.

Intrusions are “false alarms” – cases in which the participant thought they remembered something, but were incorrect. In recognition, a false alarm would be saying “yes” to an item that was not previously studied. In recall, an intrusion is an item that the participant says they remember that was not actually in the study list.

The reason both proportion correct and proportion intrusions are included for analysis is that looking at only percent correct (“hits” in signal detection theory terminology) could be misleading. If someone just said “yes” to everything in a recognition test, they would get 50% correct even if they had no memory for the items at all. You would not know they just had a bias to say “yes” unless you also looked at false alarms. If memory in one condition is really better, that condition should have both more hits (c_recog, c_recall) and fewer intrusions or false alarms (i_recog, i_recall).

You should therefore do two analyses for this lab report: One comparing the proportion correct on the two memory tests, and a second comparing the proportion of intrusions.

Steps for data analysis: (assuming you are using SPSS; you may use other analysis programs instead if you wish)

Import the data into SPSS

Save the data file to your hard drive by right-clicking the name of the file in the web browser and choosing “save as” on the pop-up menu.
Start SPSS
Open the data file in SPSS. You must select the file type of “text” and remember that the first line contains variable names.

Examine the data file for problems.

Look to see if there are missing data points, subjects who are at ceiling or floor (all correct or all incorrect), or subjects who have more than one line of data (either from repeating the experiment, or from two participants having the same number).
Look for outliers. Use the Explore option in the Analyze menu to create box plots and/or stem plots to identify outliers.
Decide what to do about outliers. Options include: eliminate that data point, eliminate that subject’s data, do nothing (keep the data as it is).

Compare the proportion correct between the two tests.

Use a t-test to compare c_recog to c_recall.
First determine what type of design is used in the experiment: Is it between-subjects or within-subjects?
Use the appropriate t-test for that type of design.
Is there a significant difference between the two conditions? Does it confirm your hypothesis?

Compare the proportion intrusions between the two tests.

Use a t-test to compare i_recog to i_recall.
First determine what type of design is used in the experiment: Is it between-subjects or within-subjects?
Use the appropriate t-test for that type of design.
Is there a significant difference between the two conditions? Does it confirm your hypothesis?
Are the results of the two t-tests consistent? In other words, do both percent correct and percent intrusions indicate that the same condition is better remembered? Think carefully about what result indicates better memory in each case.

Your lab report should be a manuscript in APA format, including all the necessary elements:

· Title Page

· Abstract

· Introduction

· Method

· Results

· Discussion

· References

· Tables and/or Figures

The Intro and Discussion can be very brief, and the references can be taken from the MEL LAB workbook’s discussion and/or the course readings list. The following will give you some pointers, and you should also consult your APA style guide.

An outline of Paper Sections, with reminders of important things to include:

Introduction

The introduction should briefly describe the topic of the experiment, previous findings from the studies this lab is based on, and the hypothesis (predicted results based on previous studies) for the lab exercises.

Method

Subjects.

Materials. Describe the stimuli in the experiment.

Design. Identify the independent and dependent variables. Identify any important counterbalancing variables. This section can be combined with the materials section if desired.

Procedure. Give a step-by-step description of what happened to the subject and what the subject did in the experiment.

Results

Describe what was done to the data. For example, what if any subjects or data points were excluded? Did you only analyze correct responses? What kind of data reduction (means per condition for each subject for example) or transformations did you do before running the analysis?

Refer the reader to a table or figure in which the summary statistics (such as means per condition) are presented. The figures and tables from your MEL LAB workbook should be used for this purpose for the lab reports. The means could also be reported in the text rather than in a table or figure. Provide a measure of variability for any means you present. (Standard deviations, standard errors, or the MSerror from an ANOVA table are all acceptable ways to report a measure of variability).

State the predicted results, or state what the question to be answered was. State the results verbally. Also report the test statistic (test of significance, with a p value) that supports your statement of the results. For example, “Participants responded faster in condition A than condition B, t (25) = 3.3, p < .05.”

Discussion

This can be very brief. Re-iterate the prediction that was based on previous studies. Summarize the results of your analysis of the lab data and state whether it confirmed the hypothesis. Briefly discuss what the result means. If the hypothesis was not confirmed, briefly discuss possible reasons that the experiment did not work.