Section 2 General Methodological Overview
2.1 Overall Methodology
The survey was released and made available to Hispanic/Latinx business owners or entrepreneurs with a business in Forsyth County, North Carolina, as well as potential Hispanic/Latinx entrepreneurs from the end of November 2022 until mid-February 2023.
2.1.1 Survey Design and Distribution
Research staff at Forsyth Futures developed the Hispanic League’s PyMEs Survey with support from the Hispanic League and the PyMEs Advisory Committee. The Hispanic League and the PyMEs Advisory Committee provided input on the survey topics and questions to ask Hispanic/Latinx business owners or entrepreneurs. Topics selected include business characteristics (e.g., number of owners, employees, etc.); information about the challenges business owners face and those planning on starting a business; and resources available or needed in the community. Business owners owning more than one business received instruction to respond from the perspective of their business with the greatest need for support.
The survey was initially developed in English by Forsyth Futures and then translated to Spanish by UNO Translations and Communications to pilot (or preview) the survey with Hispanic/Latinx business owners or entrepreneurs. A mix of English and Spanish-speaking local business owners participated in the survey pilot sessions, which provided the research team with feedback on the wording of questions (in English and Spanish), the response options, and the overall user experience to refine and improve the survey. Survey pilot sessions also allowed Spanish-speaking participants to identify mistranslations or opportunities for better phrasing in Spanish. The Hispanic League helped ensure all translations were accurate in the live survey.
After finalizing the survey instrument, the Hispanic League distributed two survey links, one in English and one in Spanish, through their social media platforms and website, the PyMEs Advisory Committee, and features in the local media. QR codes linked to the survey were shared so that participants could access the survey easily via cell phone. Incentives for survey participation included gift cards and raffles, including two tickets to the Spanish Nite Gala hosted annually by the Hispanic League. The gift card amounts were increased during the survey window to help incentivize more responses. Hispanic League staff also conducted in-person surveys with Spanish-speaking business owners via cell phone or tablet to ensure better representation in the sample survey.
2.1.2 Response Rates
Only current Hispanic/Latinx business owners, entrepreneurs, or individuals in the beginning stages of opening or planning to open a business in Forsyth County, North Carolina, were eligible to take the survey.
Data cleaning focused on ensuring those who completed the surveys were humans with a business or planning on having a business in Forsyth County. Researchers used various quality flags to identify and remove ‘internet bots’ that had spammed the survey links shared via social media, resulting in a final sample size of 129 current or prospective business owners representing more than 15 different Hispanic/Latinx heritages. Approximately 32% of respondents completed the survey in Spanish. The median time to complete the survey was 14 minutes, meaning half of the respondents took longer than 14 minutes, while the other half took less time.
2.2 Analysis Overview
Forsyth Futures’ staff members exclusively analyzed the Hispanic League’s PyMEs Survey results. The survey results presented below through visualizations and tables display the outcomes of the respondents’ answers. Key findings are regularly discussed in each sub-section to highlight notable differences and observations by Forsyth Futures’ staff to complement the visualizations and tables.
To learn more about the experiences and challenges of different business owners, analyses presented in this report share both the summary or aggregate results and disaggregated results of how different groups of people responded to the survey questions. The Hispanic League and PyMEs Advisory Committee identified the groups or disaggregations needed, while Forsyth Futures assisted in categorizing the groups. Results of the analysis are presented below through visualizations and tables; please see Data Visualizations for more detailed information on the visualizations utilized in this report.
All non-qualitative survey questions are presented in this report, however, some of the response options were collapsed or recoded during data cleaning. For questions that allowed respondents to enter their responses if theirs was not listed, analysts matched open-ended answers with existing categories. For more information on the steps taken during data cleaning, please refer to Data Cleaning in the Appendix. The data visualizations presented in the main report display the cleaned and recoded data instead of the raw or unformatted data.
The report is broken into nine sections starting with the Introduction and ending with the Appendix. The following section is Section 3, which presents the top key findings or observations from the whole report. Section 4 shows the demographic distribution of the data. Sections 5 through 8 present findings across five broadly defined categories of current owners’ business characteristics, prospective owners’ business characteristics, challenges to businesses, and resources in the community and/or at large.
2.2.1 Data Visualizations
Data visualizations throughout the report display the results of survey questions through a mix of interactive bar plots and static bar or stacked bar charts. For the interactive visualizations, please hover the mouse over the wedge or bar to see the percentages. Data visualizations presented in this report exclude respondents who did not answer the question (e.g., were missing), except when noted in specific visualizations. Please refer to the Appendix for the entire distribution of counts and percentages by survey question and disaggregation.
In the Business Characteristics: Current Owners, Business Characteristics: Prospective Owners, Challenges to Businesses, and Resources sections, the static bar plots and stacked bar charts display proportions by answer choice for the survey questions and disaggregations that were determined to be statistically significant. Disaggregations were determined to be statistically significant if the tests analysts ran indicated that the relationship between the disaggregation and question responses was stronger than what could be attributed to random chance. For example, suppose a visualization includes information on how respondents of different genders answered a question. In this case, statistical tests indicated that the relationship between the gender of the respondent and how they answered the question was strong enough that it was not likely the result of random chance.
When data is disaggregated, it is broken into smaller groups to answer a question. Disaggregating data does not necessarily mean that any differences between groups are strong enough to result from anything other than random chance. In other words, just because disaggregated data shows a pattern does not necessarily mean it is not a coincidence. For example, a significant relationship between gender and responses to a question does not mean that an apparent difference between how men and women answered a question is significant. The evident difference in responses could still be the result of random chance. There are some exceptions to this general rule, and the cases where analysts can be confident that significant individual differences exist are noted in the analysis text.
It is also important to note that a relationship not being identified as significant does not mean it does not exist but that there is insufficient evidence in the available data that a relationship exists. Variation in how consistently people give similar answers, the number of people in a disaggregation group, and the number of response options can all influence the likelihood of enough evidence to be confident that a relationship exists. Relationships are particularly likely to have little findings in cases where there are many response options or a small number of people in one of the groups in the disaggregation.