Supporting Statement B

B. Supporting Statement B
This part of the proposed information collection involves statistical methods.
B1. Respondent Universe and Sampling Methods
The study population for this effort will be made up of individuals age 18 and over.
Respondents will represent a mix of demographics. However, the study targets the following:
female (50%), male (50%); Midwest (~20%), Northeast (~20%), South (~40%), West (~20%).
Ethnicity of participants will be monitored in line with the following criteria: Non-Hispanic
White (~66%), Non-Hispanic Black (~12%), Hispanic (~12%), and Other (~10%). The panel
provider will monitor respondents, and if a particular demographic is trending highly, the panel
provider will slow down the sample for that segment and will focus on obtaining responses
from others to ensure recruitment for U.S. census-matched survey participants from the
Midwest, Northeast, South, and West regions. Certain segments of the sample, including
underserved populations, will be included. The survey panel will monitor and screen to ensure
that insights are collected from a diverse population.
For this survey, a form of non-probability sampling, called quota sampling, will be used. Quota
sampling offers a way to survey individuals who are easy to reach; however, this sampling
method is not guaranteed to produce a representative sample. Accordingly, findings from the
survey may not be valid to generalize to the overall population.

B2. Procedures for the Collection of Information
Contractor will partner with one online sample provider. Although the sample will not be
probability-based, the Contractor’s partner will randomly select respondents for surveys where
respondents are likely to qualify. Respondents are invited to survey in various ways. Often,
potential respondents are sent an email invitation informing them that the survey is for research
purposes only, detailing how long the survey is expected to take, and explaining what
incentives are available. Not every member of the population has an equal chance of being
included in the sample. Members may unsubscribe at any time. Other times, respondents will
see surveys they are likely to qualify for, upon signing into a panel portal. Other common
invitation methods include in-app notifications and SMS notifications. Therefore, procedures
are not guaranteed to produce a representative sample. To avoid self-selection bias, survey
invitations do not include specific details about the contents of the survey, and instead, the
invitations are kept very general. However, various forms of bias may not be avoided.
B3. Methods to Maximize Response Rates, Deal with Bias, and No Response
The panel provider will field the survey until it obtains the minimum 4,000 respondents with
a goal of 5,000 completes. The panel provider will monitor sample performance during the
field period and regularly update regarding completion status. During the fielding period, the
Supporting Statement B, Page 3 of 4

panel provider will send email reminders to non-responders and take other actions, as needed,
to ensure a mix of demographics.
CPSC staff does not believe survey results will be affected negatively by the Covid-19
pandemic. Although this survey tries to make connections between purchasing factors online
and purchasing factors in-person, for various consumer products, the timing of the proposed
survey, in relationship to the pandemic, will provide a point-in-time estimate. “In-person”
purchasing habits may not return to pre-pandemic levels. Regardless, any results for rates of
online shopping will not be generalized to the population because online survey respondents
may be more likely to shop online.
B4. Tests of Procedures or Methods to Be Undertaken
A test panel of seven initial respondents showed that the survey might take an average of 6
minutes to complete, with the maximum duration for the survey response lasting 8 minutes.
Additional standard pre-testing procedures to be undertaken by the panel provider will include
inviting another small number of respondents to participate in the survey as case studies, to
ensure that the survey is of adequate length and is processing correctly for respondents that
will reflect the potential panel. If any changes need to be made, the panel provider will do so,
accordingly, and will proceed with fielding the survey to the rest of the panel.
B5. Individuals Consulted on Statistical Aspects and Individuals Collecting and/or
Analyzing Information
CPSC staff and Contractor staff identified below, extensively developed and reviewed the
proposed protocol and survey. CPSC and Contractor staff will analyze the information and
create technical reports.
Rana Balci-Sinha
Director, Division of Human Factors
U.S. Consumer Product Safety Commission
5 Research Place, Rockville MD 20850
Phone: 301-987-2584
[email protected]
Michaela Beckenbach
Principal Research Manager
Research Services Operations at Qualtrics
[email protected]
Bill McDonald
Qualtrics at Carahsoft
11493 Sunset Hills Road
Reston, Virginia 20190
Supporting Statement B, Page 4 of 4