Newsletters

Economic Emergency Information Collection Question Bank: a quick introduction on how to adapt the questions for your establishment survey

01/11/2024

Katherine Blackburn, RTI

Y. Patrick Hsieh, PhD, RTI

Chris Ellis, RTI

In March 2022, the Census Bureau, with support from RTI International, published the first edition of the Emergency Economic Information Collection (EEIC) Question Bank, consisting of a total of 173 unique questions, accessible at https://tinyurl.com/EEICQBank. The Question Bank is the culmination of a year of work to develop a large set of “at-the-ready” cognitively tested survey questions to measure the economic impact of emergencies, especially natural disasters, on U.S. establishments. The Question Bank includes topics such as employees and payroll, revenue and expenses, supply chain, closures, and work location, as well as some sector-specific questions for manufacturing, retail, and non-profit industries. Work has recently concluded on the EEIC Question Bank v2.0, which will be available publicly in the near future and expands on the first edition, including additional questions targeting the public sector. Federal government agencies may apply for OMB generic clearance for their rapid data collection needs after emergencies with questions sourced from the EEIC Question Bank.

The questions featured in the publicly accessible Question Bank (https://tinyurl.com/EEICQBank) were all cognitively tested with a minimum of 15 participants from the relevant sectors and have been deemed methodologically sound for data collection. The Question Bank includes cognitive testing notes and design recommendations and serves as a valuable resource for establishment survey designers on a variety of topics beyond the context of emergencies. For example, one popular topic included in the Question Bank is remote work, which continues to evolve after the initial impacts of the Covid-19 pandemic. The remainder of this article will walk through how to search the Question Bank for relevant questions (using prevalence of remote work as an example) and review the information provided to decide which question to employ.

The first place to start with the extensive Question Bank document is the Table of Contents, which outlines all of the topics included in the Question Bank. From the Table of Contents (Exhibit 1) we can see that the topic of remote work is included under Closures and Work Location (Section 11).

 

 

 

 

 

 

 

 

 

 

 

 

After navigating to the section for changes in work location/remote work, we see the first question, Q113 (Exhibit 2). Survey designers should review potential gate questions, alternate and complementary questions, and other notes for each question. From these pieces of information for Q113, we surmise that this question has high respondent burden due to asking for a percentage of remote workers, for which companies often do not keep records.

Exhibit 2. Remote Work Question Q113
Q # Q113
Topic tags Changes in work location / remote work
Question Wording <Reference period>, what percent of all employees at this <business/agency/etc.> worked from home at the following frequencies?

Enter 0 if none. Estimates are acceptable.

If grid, subitems Never

One day per week

Two to four days per week

Five to seven days per week

Total [PROGRAMMER: TOTAL COLUMN]

Response options [Numeric open-end] %
Sector applicability
Gate question/Follow-up design consideration Recommended skip pattern: Q128 is the follow-up to Q113.

Optional skip pattern: Can use Q142 as the follow-up to Q113.

Question type Open-end numeric
Alternate/Complementary questions Q127 asks how the number of employees working from home one day or more per week changed due to the event. Q127 is not as burdensome for respondents to answer, as it does not ask for exact percentages.

Q141 asks about the percentage of employees that will work from home in a future time period.

Other questions that ask about employees working from home: Q56_D, Q128, Q139, Q142, Q206

Reference period fill From <date 1> to <date 2>

In the <month 1> — <month 2> quarter of <year>

In <month> <year>

Since <date>

Reference period notes Reference period should be a period of time/date after event.

If using “Since <date>”, date should be date of event

Other notes This question does not ask about the direct impact of the event. Users should consider asking this question before any question that asks about the direct impact of the event to prevent question context effects (see user manual for details).

Testing indicated that this question is burdensome for respondents to answer. If respondents have records that allow them to answer, it requires time and effort to find these numbers. If respondents don’t have records that allow them to answer, they would have to consult with others and may have to estimate/make a guess.

 

From the alternate/complementary questions information provided on Q113 we can see that Q127 is noted as having less respondent burden, so we could consider that question next, presented in Exhibit 3. Q127 has a lower respondent burden because it asks for general trends instead of precise numeric data.

Exhibit 3. Remote Work Question Q127
Q # Q127
Topic tags Changes in work location / remote work
Question Wording Compared to what was normal before <event>, how has the number of employees working from home one day or more per week changed as a result of <event>?
If grid, subitems
Response options Increased

Did not change

Decreased

Not applicable

Sector applicability
Gate question/Follow-up design consideration
Question type Select one answer
Alternate/Complementary questions Q113 asks what percentage of all employees work from home at several frequencies. However, Q113 is more burdensome for respondents to answer than Q127 and Q113 does not ask about the direct impact of the event.

Other questions that ask about employees working from home: Q56_D, Q128, Q139, Q141, Q142, Q206

Reference period fill
Reference period notes
Other notes

 

Another distinction of Q127 is that it asks for a comparison to what was normal before an event, which may or may not be useful for a given survey. During the additional testing for the EEIC Question Bank v2.0 Q127 was also tested with the public sector. The second edition of the Question Bank will include additional sector-specific notes for Q127 (Exhibit 4) and other questions that were tested with the public sector.

Exhibit 4. Sector Applicability Notes for Q127 in EEIC Question Bank v2.0
Sector applicability Private Sector:

·       Applicable cross-sector.

·       This question wording does not include the <employee type> fill. At the time of testing, respondents indicated that their companies did not often have records about employees who work from home. Therefore, it would be too burdensome to ask this question about specific kinds of employees.

Public Sector Testing Notes:

·       RTI tested this question with the public sector and testing supported using this question with the public sector.

·       This question wording does not include the <employee type> fill. At the time of testing, respondents indicated that their companies did not often have records about employees who work from home. Therefore, it would be too burdensome to ask this question about specific kinds of employees.

·       However, it may be important to specify “paid employees” when using with the public sector.

·       Note that when testing with the public sector, this question used the wording “paid employees.” Respondents used “Did not change” and “Not applicable” interchangeably at times to mean there were never any employees who worked from home, or that employees were never allowed to work from home. Additionally, the “not applicable” response option is used by respondents when the agency is only staffed with volunteers.

 

Using the information provided for each of these questions, we can consider which one may best fit our analytic needs. If neither question fits the exact needs of the survey, the notes still inform potential development of a new question (e.g., establishments typically do not have records about remote work and that asking for precise numeric data carries higher respondent burden).