PMBC Data Set Publications

We have conducted replication analyses of three of the publications that derived from PMBC data:

1. Janicki-Deverts, D., Zilles, K., Cohen, S., & Baum, A. (2006). Can a 15-hour (overnight) urinary catecholamine measure substitute for a 24-hour measure?  J Appl Biobehav Res, 11, 1-11 (variable list 1).
2. Cohen, S. & Lemay, E. (2007).  Why would social networks be linked to affect and health practices? Health Psychology, 26, 410-417 (variable list 2).
3. Cohen, S., Alper, C.M., Doyle, W.J., Adler, N., Treanor, J.J., & Turner, R.B.  (2008). Objective and subjective socioeconomic status and susceptibility to the common cold.  Health Psychology, 27, 268-274 (variable list 3).

Inconsistencies between results from these publications and the PMBC data replications are described in detail below.  For additional information on the data set variables that were used in each analysis, click on the links indicated above.

Inconsistencies Between Results Published in Janicki-Deverts, et al (2006, J Appl Biobehav Res) and PMBC Data Set Replication.

The analyses reported in Janicki-Deverts et al (2006) used total catecholamine production rather than catecholamine concentration in urine.  Results of the main analyses substituting the concentration values rather than total production are provided below. 

Table 1. Correlations of 15–Hr and 24–Hr Urinary Catecholamine Measures

(a) Untransformed Data

***p < .001

Relative efficiencies:

Epinephrine: 72.3%

Norepinephrine: 82.8%

(b) Log₁₀-transformed Data

***p < .001

Relative efficiencies:

Epinephrine:  70.6%

Norepinephrine:  79.2%

Table 2.  Correlations of Urinary Epinephrine & Norepinephrine with Blood Pressure & Salivary Cortisol‡

†p < .10. *p < .05. **p < .01. ***p < .001.

‡ For analyses by sex and race (manuscript tables 3 and 4), see the attached file.

Inconsistencies Between Results Published in Cohen et al (2007, Health Psychology) and PMBC Data Set Replication

Infection

The Methods section of the paper reports that 152 participants had been infected with the cold virus.  This number should be 155.

Objective Colds. 

The criterion used to determine whether participants had developed an objective cold differed from the criterion used in all Common Cold Project data sets:  infection + [(mucus weight > 10 g) or (nasal clearance time > 35 min)].  Employing a greater than criterion rather than a greater than or equal to criterion resulted in 2 participants not being classified as having a cold.

Inconsistencies Between Results Published in Cohen et al (2008, Health Psychology) and PMBC Data Set Replication

Objective Colds

More stringent criterion was used (see above comments on Cohen et al., 2007)

One participant who was missing baseline mucus weight data was assigned a positive objective cold score.  In the Common Cold Project data sets, this participant’s cold score is missing.

Subjective Symptoms

The subjective symptom score computed for Cohen et al (2008) was comprised of cold, flu, and cold/flu complication symptoms rather than the 8 Jackson symptoms used in the Common Cold Project data sets.  Specifically, total upper respiratory symptoms for each day were calculated as the sum of the following:

• Nasal congestion
• Sneezing
• Runny nose
• Ear ache
• Sinus pain
• Sore throat
• Cough
• Chest congestion

Interview alcohol consumption

Participants with missing data for the indicator (i.e., “did you consume alcohol today?”) were assigned a 0 score for the quantity variable rather than a missing value.

Extraversion

Score were computed from the average of 2 administrations, with one being rated on a 1-5 scale and the other on a 0-4 scale.  No adjustment was made for this difference in scaling.