Public Health Emergency Preparedness

2.4.  Interpreting the Analysis File

2.4.1.  Background

After viewing the data in MS Excel or importing into MS Access or SPSS, the analysis will have a file with multiple columns. Each column corresponds to a specific piece of information collected by the questionnaire. Before we explain how to interpret that file, to ensure a common usage of terminology in this section, the following definitions are provided.

Analyst: This is the individual responsible for collating and analyzing the questionnaire results from respondents in his/her jurisdiction.

Analysis file: The file that the response data is imported into. This will either be MS Access, MS Excel, or a statistical package such as SPSS. Refer to Section 2.3, "How to Download and Import Data," for a detailed description of the steps to export the .csv or .txt file into one of the appropriate tools. 

Mapping document: The document that provides a legend or a key to translating the analysis file. It will provide the means by which to understand the analysis file. This file is included as Appendix B.

Respondent: The individual completing the questionnaire.

Parent question: The main question as it appears in the questionnaire.

Response scale: The potential answers to the parent question. All 43 questions have a standard response scale, and they vary slightly depending on the question. In general, the responses range from "No, and we don't plan on performing (the activity)," to "No, but we plan on performing (the activity)," to "(the activity) is in progress," to "Yes, but there are limitations to our ability to complete (the activity)," to "Yes, and we exceed" (the activity).

Followup questions: Any additional questions that appear if a respondent answers the parent question in a certain manner. These primarily appear when the respondent chooses one of the response options that begins with "Yes, and..." or "Yes, but..."

The following illustrates an example of a parent question, its response scale, and the followup questions:

Parent question:

3  Does the hospital use an Incident Command System (ICS) to manage events that impact normal operations?

Response scale: 

___  Response #1: No, and not planned within the next 6 months.
___  Response #2: No, but the hospital plans to use an ICS within the next 6 months.
___  Response #3: ICS is currently being developed.
___  Response #4: Yes, but all hospital staff are not trained on their roles in the system.
___  Response #5: Yes, and all hospital staff are trained on their roles in the system.
___  Response #6: Other. 

Followup questions:

(Table will be activated if Response #4 or Response #5 is selected.)

Select the appropriate response for each National Incident Management System (NIMS) activity.	Y or N
Is the ICS used on a near daily basis to manage events that impact normal operations?	Y N
Is the ICS practiced routinely in exercises/drills?	Y N
Is the ICS updated as needed after exercises/drills?	Y N
Is the ICS incorporated into existing training programs?	Y N
Is the ICS formally incorporated into the emergency operations plan (EOP)?	Y N
Is the ICS coordinated with local entities?	Y N

For the remainder of this section, we recommend having a sample analysis file, a blank version of the questionnaire, and the file titled, "final_mapping document.xls" at hand. Having sample documents accessible will help make it easier to follow and conceptualize the materials presented below.

Return to Contents

2.4.2.  Question Numbering Scheme

Though there are 43 questions in the questionnaire, through the use of the followup questions, there is the potential for up to 316 questions asked in the questionnaire. For example, for question 3 above, if the respondent answered either of the "yes" responses, the table would activate with six more questions for the respondent to answer, for a total of seven questions, all associated with question 3. This includes the parent question and six followup questions. 

In the data analysis file, each parent and followup question is represented by a column. Each row corresponds to one respondent who completed the questionnaire. The following figure depicts an example taken from an Access analysis file.

Because a textual description of the question represented by each column would make the analysis file large and cumbersome, we present a numeric scheme for uniquely associating each column in the analysis file to a parent or followup question on the questionnaire. In the following sections, we will explain this numeric system. 

There are three sections to the analysis file. These include the columns that contain the respondent's unique identifiers, the demographic questions, and the survey content questions.

Unique Identifiers:

The analysis file will begin with column headers that provide the Survey ID number and the begin and end times. These are included for tracking purposes, to ensure that the analyst can uniquely identify each respondent that completed the questionnaire. The following figure depicts a sample demographic portion of an analysis file in MS Access. 

These fields are defined below:

SurveyID: A unique identifier assigned by the system to each respondent who completes the questionnaire. 

BeginTime: The date and time the respondent began entering data into the Web tool.

EndTime: The date and time the respondent stopped entering data into the Web tool.

The "BeginTime" and "EndTime" columns are not critical for the analysis of the data but are provided as additional data points for tracking respondents.

Demographic Questions: Following the unique identifiers are the answers to the demographic questions, noted as D1 through D21. The first seven demographic questions are shown in the figure below.

To determine exactly what survey question is represented by each column, the analyst can refer to the mapping document in Appendix B.

The mapping document matches each column header in the analysis file to its corresponding question. For example, D01 corresponds to the question that asks the hospital name; D02 refers to the second demographic question, "Street address"; D03 corresponds to the third demographic question, "City"; etc. In the mapping document, you will see that each demographic question has a column ID associated with it as well as a text-based description of the questions and answers.

Survey Content Questions: The figure below illustrates the columns that correspond to the survey content questions, designated with a "Q." 

Note: For reasons of space restriction, this file shows only the Survey ID and the beginning of the survey questions. In an actual analysis file, all of the unique identifiers and demographic data would still be present and would appear in the columns before the start of the survey content questions.

The survey content questions are coded in the same manner as the questions in the demographic section. Parent questions are Q03, Q04, etc. If parent questions have followup questions, those have extended codes (Q034A, Q034B, etc.). We will describe this numbering scheme below.

From the example question presented at the beginning of the section (question 3), there is a six point response scale. 

Response scale: 

___  Response #1:  No, and not planned within the next 6 months.
___  Response #2:  No, but the hospital plans to use an ICS within the next 6 months.
___  Response #3:  ICS is currently being developed.
___  Response #4: Yes, but all hospital staff are not trained on their roles in the system.
___  Response #5: Yes, and all hospital staff are trained on their roles in the system.
___  Response #6: Other.

In the analysis file, for the followup questions, following the parent question (Q03), the next digit will be a "4" to indicate that the user chose either response 4 or 5. Please note that to make the naming convention more consistent, even if the respondent selected response 5, the column will still utilize a "4" in the naming convention. Then, the associated followup questions are labeled with an alpha character. We will use column ID "Q034A" as an example.  If a respondent chose response 4, "Yes, but all hospital staff are not trained on their roles in the system" or response 5, "Yes, and all hospital staff are trained on their roles in the system," the system presented the respondent with a table with additional information requests. The additional table for question 03 is as follows: 

(Table will be activated if Response #4 or Response #5 is selected.)

Select the appropriate response for each National Incident Management System (NIMS) activity.	Y or N
Is the ICS used on a near daily basis to manage events that impact normal operations?	Y N
Is the ICS practiced routinely in exercises/drills?	Y N
Is the ICS updated as needed after exercises/drills?	Y N
Is the ICS incorporated into existing training programs?	Y N
Is the ICS formally incorporated into the emergency operations plan (EOP)?	Y N
Is the ICS coordinated with local entities?	Y N

Therefore, the first question in the table, "Is the ICS used on a near daily basis to manage events that impact normal operations?" would have a column ID of Q034A in the analysis file. Q034B corresponds to "Is the ICS practiced routinely in exercises/drills?" etc. This is further explained in the figure below:

Return to Contents
Proceed to Next Section