Attachment H1_SamplingLargeOrgs

Attachment H1_SamplingLargeOrgs_FINAL_08042010.docx

Survey of Healthcare Workers' Health and Safety Practices

Attachment H1_SamplingLargeOrgs

OMB: 0920-0860

Document [docx]
Download: docx | pdf






Attachment H. NIOSH Sampling Approach for Partnering Organizations



Attachment H1: Organizations with greater than 3,000 members with email addresses


Simple random sample


In order to conduct a survey of your members, your organization will need to prepare a targeted sampling frame. This will involve the following steps: 1) generate a list of members with email addresses; 2) filter out members such as retirees, students, administrators/directors/managers, educators, and others who are not likely exposed to [insert relevant chemical agent(s)], 3) retain members who are likely to use or come in contact with these chemical agents, and 4) draw a random sample from the filtered/targeted membership population. A quality check of the sampling frame needs to be performed to ensure that excluded members are not in the sample.


Our goal is to obtain 1,000 completed surveys (i.e., hazard module and core module). Assuming a 50% response rate, a sample size of 2,000 will be necessary to ensure that we meet our goal.


Step-by-step instructions on how to select a random sample from a list of individuals (using MS Excel) is provided. NIOSH can provide assistance to your staff with this process, if requested.


The sample will need to be drawn within two months of launching the survey to verify that the sample is current. The survey is scheduled to be launched in late January 2011.


For quality assurance purposes, a written description of the steps taken to prepare the sampling frame will need to be provided to NIOSH.




Instructions for sampling membership (for Organizations with greater than 3,000 members)

For this method, your membership data will need to be in MS Excel. In this example, we are using 30 made-up names and email addresses.

This represents our “targeted sampling frame” as all retirees, students, administrators/directors/managers, etc have been filtered out.

Our final sample size will be 10 people.



Step 1 creating random numbers.

In the next column (column D) type in “random#”. In cell row 2, column D type in the formula “=RAND()” It will show up as a number in the actual table.


Shape2 Shape1

Step 2.

Copy and paste “=rand()” into all the cells in the column across from organization members.


Shape3

Step 3

Now, highlight column D with the random numbers in it and go to “Edit” and copy the column.


Shape4


Step 4

Step 4.

Next highlight the next column (column E) and go to the Edit menu and highlight “Paste Special”


Shape5

Step 5

Click on the button that says “values”, then click “ok”


Shape7 Shape6

Column E will contain numbers, not a formula. Column D will continue to generate random numbers.


Shape8

Step 6. Highlight the column with the formula (column D)


Shape9

and delete it (EditDelete)


Shape10 Step 7.

Highlight all cells in spreadsheet by clicking the square in the left hand corner of the spreadsheet.


Shape11


Step 8. Choose Data, then Sort

Shape12

Step 7.

Sort by “random#” (column D). With the “Ascending” button chosen. Click “Ok”.


Shape13 Shape14 Step 8.

Note that in column “D” numbers are in order from smallest to largest. Since our sample size for this exercise was 10, the first 10 people (rows 2 to 11) in the list are now our sample and the others can be deleted.


File Typeapplication/vnd.openxmlformats-officedocument.wordprocessingml.document
File TitleNIOSH Sampling Approach for Partnering Organizations
Authorjmb4
File Modified0000-00-00
File Created2021-02-02

© 2024 OMB.report | Privacy Policy