
Strategic Blending in the Age of AI: Why Automation Alone Isn’t Enough
October 9, 2025
The Big Qualtrics Acquisition, Melanie’s New Role, and Halloween Candy
October 22, 2025One of the biggest topics in the news right now is the deployment of the National Guard to various U.S. cities to assist with immigration enforcement and crime. These actions are very polarizing for many Americans. We wanted to better understand in what situations people would support the deployment of the National Guard, so we asked in our most recent round of research-on-research.
Let’s dive into what we uncovered.
Overall
Across all respondents, 58% support National Guard deployment during natural disasters, making it the most widely accepted scenario. Major public safety outages, such as widespread power or communication failures, follow at 47%, while border security (42%) and protest or crowd control (35%) receive less support. Only 30% favor deployment during routine crime surges, and 15% oppose all these situations entirely.

Gender
Gender differences among respondents were modest but noticeable. Women showed slightly higher support for deployment in natural disasters (60% vs. 56%) and were more likely than men to select “none of the above” (17% vs. 12%), indicating a stronger preference for limiting National Guard use. Men expressed more support in scenarios involving border security (45% vs. 39%) and public safety outages (49% vs. 45%), showing more willingness to use the Guard for security-related issues.

Age
Age presented one of the clearest trends in the data. Support for National Guard involvement increased steadily with age, from 42% among those 18–24 to 83% among those 65 and older for natural disasters. The same pattern appeared across other categories, including border security and public safety outages. Older Americans were markedly more supportive of Guard deployment across nearly all contexts, reflecting stronger trust in formal institutions. Younger respondents tended to be more cautious, with one in five 18–24-year-olds saying none of the scenarios justified deployment.

Income
When looking at the data by income, there was a similar pattern to the split by age. Higher-income respondents consistently expressed stronger support for National Guard action, particularly for major outages (58% among those earning $100,000+) and border security (54%). Lower-income respondents were less likely to favor deployment and more likely to oppose all listed scenarios (23% among those under $20,000).

Political Affiliation
Unsurprisingly, political affiliation marked the sharpest divide in opinions. Republicans were significantly more supportive of deploying the Guard across nearly all situations, especially for border security (63%) and crowd control (53%). Democrats showed higher agreement for natural disasters (61%) but considerably lower for border security (29%) and protests (25%). Independents generally fell between the two, while those identifying as “Other” expressed the lowest overall support and the highest rate of selecting “none of the above” (37%).

Ethnicity
When we broke the data down by ethnicity, there were some trends that emerged. Caucasian respondents reported the highest support across most categories, particularly for natural disasters (63%) and border security (46%). African American respondents showed lower agreement across nearly every situation, with only 43% supporting deployment for natural disasters and 31% for border security. Asian/South Asian and Hispanic/Latino respondents generally aligned in the middle, both more supportive than African Americans but less so than Caucasians.

Panel
Finally, looking at the data by panel source showed some distinct variation in levels of support across the different situations. Panel M, for example, recorded higher approval rates for natural disasters (79%), while Panel U reported lower figures (47%). Others, like Panels S and D, leaned toward broader support in crisis situations.

These results highlight why strategically blending your sample is so important for online quantitative studies. By not doing so, you could end up with biased data that could lead your client to the wrong business decision. To learn more about strategic sample blending, click the button below.



