 |
 |
 |
 |
 |
 |
 |
Survey home > Our Methodology
Our Methodology
In February of 2008, Today’s Hospitalist, a Roman Press publication, commissioned Accelara Research to conduct research among its Hospitalist audience to develop a current picture of employment and compensation conditions throughout the industry.
The survey was fielded between May 7th and July 7th, 2008. Respondents were invited to the survey using three methods:
- A full-page ad was placed in the June 2008 issue, which arrived beginning May 7th, inviting readers to an online survey hosted www.todayshospitalist.com. Ad respondents were offered an opportunity to enter a drawing for $500.
- An email invitation was sent by Today’s Hospitalist to approximately 4,000 (n=3,984) readers with email addresses on file on May 27th, and again on June 18th. Email respondents were offered an opportunity to enter a drawing for $500 (along with ad respondents).
- A direct mail paper survey was mailed to n=1000 respondents in Today’s Hospitalist envelopes, with a personalized cover letter from publisher Edward Doyle requesting participation. Respondents had received a post-card invitation several days before the mailout, and they received a reminder invitation 7 to 10 days afterwards. A $10 personal check was written to each respondent as incentive to participate in the survey. Respondents were invited to respond either on paper or online. Survey packets were mailed on June 3rd and 4th.
Of the total 682 responses, 162 arrived on paper, 570 took the online survey.
In total, 682 responses were received, of which 675 were sufficiently complete and arrived in time for tabulation. The majority of respondents arrived from either the e-mail or the direct mail solicitation.
The statistical confidence interval at a 95% level of confidence* for this sample is +/- 3.8% for statistics near the 50%/50% range, and +/- 2.3% for statistics at a 90%/10% range.
This is the first comprehensive independent survey to look at the compensation of individual hospitalists. From a statistical perspective, this survey differs from surveys conducted by societies because it collected information directly from hospitalists, not hospitalist practice managers.
Salary averages may be understated due to truncating high-end data at too low a point ($250K) and using a high-end average of only $254K, while the actual average may be much higher.
At the same time, we asked for salaries within specified ranges and then used the midpoint of the range for our calculation of averages, which may introduce a slight skew upwards if the actual earnings are skewed toward the lower end within each category (This would only account for up to $5K, however, since the compensation brackets were fairly tight.)
*Where 19 of 20 random samples are likely to show results within this range). Statistical confidence is a function of final sample size.
**Note that in some instances figures add up to more than 100% because of multiple selections. In other instances figures do not add up to 100% because of rounding errors.
