The experiment was designed to examine if the control group (no stressor) differed from the experimental group (stressor) in object-location memory score due to cortisol level. A 2x5 ANOVA was performed 2 (group= control, experimental) x 5 (cortisol measurements).  A main effect for cortisol level was found F(4, 44)= 3.611, P=0.012.  Within subject analyses examined differences across manipulation in cortisol level.  There was no group (experimental versus control) by cortisol interaction F(4, 44)=0.522, P=0.72.  Planned comparisons to examine cortisol level on a measurement-by-measurement basis were calculated with the appropriate error term from the preceding ANOVA using the Fisher’s LSD procedure.  Upon comparison, significant differences were found between cortisol measurement 1 and cortisol 4, cortisol 1 and cortisol 5, cortisol 2 and cortisol 4, and cortisol 3 and cortisol 4.  No significant differences F(1,11)= 0.036, P=0.854  were found between subjects for the performed 2 x 5 ANOVA examining differences in cortisol level by group.  To examine if there were any differences between the control and experimental groups in their baseline cortisol levels, an independent samples t-test was performed and found no significant differences t(11)= -0.231, P=0.822.
         An ANOVA for object-location memory was also examined.  There was a significant main effect for object-location memory scores at time 1 and time 2 when a 2 (group=control, experimental) x 2 (trial 1 of object-location memory test, trial 2 of object-location memory test) ANOVA was performed F (1, 11)=6.75, P=0.025.   No interaction between group and test were found for within group comparison F (1, 11)=2.408, P=0.149.  Between subject comparisons revealed no significant differences between the control and experimental group on either trial of the object-location memory test F (1, 11)=0.515, P=0.488.  Even though the differences were not significant, the experimental group (M=8.50, SD=3.02) had slightly lower scores than the control group (M=10.43, SD=2.23) on the first trial of the object-location memory task and marginally higher scores on the second trial (M=11.33, SD= 1.75, M=11.14, SD= 2.79).
        After examining means, correlational analyses were performed to examine the relationship between the change in cortisol and test score (see Table 1 for control group, Table 2 for experimental group).  Correlational analyses found significant relationship between different changes in cortisol level and certain object-location memory scores for both the control and experimental group. Significant Pearson correlations were found for the control group between the number of correct answers on the object-location memory test (trial 1) and the change in cortisol level from measurement 1 (baseline) to measurement 2 (after 5 minute reading period; r=-0.787, P=0.018, one-tailed).  The change from cortisol 1 (baseline) to cortisol 4 (after second 5 minute reading period) was significantly correlated with the number of correct answers on trial 1 of the object-location memory test for the control group as well (r=-0.681, P=0.046).
             In the experimental group, the number of wrong hits (e.g., incorrect answers) on the object location memory test (trial 1) related significantly to the overall change in cortisol level from measurement 1 (baseline; r=-0.729, P=0.050, one-tailed).  The change of cortisol from baseline to measurement 4 (after second trial of stressor) was significantly correlated with wrong hits in the first and second trials of the object-location memory test (trial 1, r=-0.819, P=0.023; trial 2, -0.730, P=0.050).  Wrong hits in trial 1 was also significantly associated with  the change from measurement 3 to measurement 4 (r=-0.899, P=0.007).  In trial 2 of the object-location memory test, wrong hits were significantly correlated with the change in cortisol from baseline to measurement 3 (r=-0.873, P=0.012; see Table 2 for complete summary of correlational analyses).  The cortisol measurement in the closest temporal space (cortisol 5) was not significantly correlated with the Self-Rate Stress questionnaire for the control group (r=0.271, P= 0.278, one tailed) or for the experimental group (r=-0.109, P=0.419).
         In performing exploratory analyses, individual trends for cortisol level across time were plotted (see Figures 3-10).  Three of the control subjects exhibited increases in cortisol level (change= 1.95 nanograms per milliliter (ng/ml), 0.27 ng/ml, 0.04 ng/ml; see figures 3-5) after performing the first trial of the object-location memory task. One control subject showed an increase in cortisol level (change= 4.08 ng/ml see figure 6) after the second trial of the object-location memory task as well (see figure 4).  Conversely, only one experimental subject showed similar trends (change= 0.46 ng/ml).  Two experimental subject showed reactivity to the stressor with an increase in cortisol level from baseline (change= 0.23 ng/ml, 0.18 ng/ml; see figures 7 and 8).  The overall trend in change was a decrease in cortisol level in time (control: M=-0.47, SD=0.67; experimental: M=-0.93, SD=1.08).  Four out of the six experimental subjects had a decrease in cortisol after performing the first trial of the object-location memory test (change= -0.30 ng/ml, -0.63 ng/ml, -0.55 ng/ml, -0.21 ng/ml; see figures 7-10).  These results were not significant and were plotted for qualitative purposes.
        Object-location memory scores were converted to z-scores for qualitative comparison to changes in cortisol at certain points in the manipulation.  Specific subjects results, showing interesting patterns, were examined (see figures 4, 6-10).  Two subjects had a slight increase in cortisol at various time points, and compared to other subjects, had moderate cortisol levels (range 6-9 ng/ml; see figure 4, 6).  One subject had a decrease in cortisol, but still maintained moderate levels of cortisol (6-8 ng/ml; see figure 8).  For the two subjects with moderate cortisol levels that increased (range 6-9 ng/ml), their z-scores for object-location memory were above the mean.  Subject #7 had z scores of 0.67 and 1.15 while subject #2 had z-scores of 0.31 and 0.26.  Subject #8 whose cortisol level remained moderate, but decreased had z- scores both above and below the mean (z= 0.31, -0.64)  The three subjects with low levels (1-2 ng/ml) of cortisol when compared to other subjects had z-scores that fell below the mean (subject # 12 z= -0.05, -1.99; subject # 14 z= -0.77, -0.64; subject #4 z= -0.41, -2.88).
 
 

click here to return to main page

(to discussion)