Why are the trend data for F3 Marketing and Recruiting Budgets in the 2024 EMBAC Program Survey report different from the data reported in 2023?

This is the question F3 from the 2024 EMBAC Program Survey:

F3. Please estimate the percentage of your marketing and recruiting budget that was allocated to the following categories in your most recent or current fiscal year. Please enter whole numbers and do not enter commas or a % symbol. The sum must add up to 100%.

	% of Marketing & Recruiting Budget
Digital and social media marketing	[NUMERICAL BOX]
Operating Expenses not including faculty costs (residencies, food, books, etc.) – not including staff salaries	[NUMERICAL BOX]
Advertising via TV, radio, newspapers, & magazines	[NUMERICAL BOX]
In-person information sessions	[NUMERICAL BOX]
Virtual information sessions (e.g., webinars, live meetings, etc.)	[NUMERICAL BOX]
Other	[NUMERICAL BOX]
Total	[NUMERICAL BOX]

Improved Question-Level Data Provisioning

In 2024, Percept Research transitioned data analysis to Power BI for validating question-level responses. This ensures that only schools responding to each specific metric are included in the calculated variables (e.g., Acceptance Rate, Enrollment Yield). This approach enhances the accuracy of metrics and strengthens the validity of trends.

Due to this transition, the methodology for handling ‘blank’ responses was adjusted. Blank responses were transformed to '0' where the total result across all categories = 100% (this is footnoted on the pages where F3 is visualized in the EMBAC Program Survey reports).

Why was this adjusted?

Each segment in a stacked bar chart represents a portion of the whole. Blank responses can distort this proportionality, making it harder to interpret the data accurately. Assigning '0' ensures the visual representation is precise and each segment's size correctly reflects its value.

Previously, for example, a variable (such as ‘Advertising via TV, radio, newspapers, & magazines’) would report an average of 15 for three school responses: 10, 20, and 'blank'. This results in an overflation of the lesser-used marketing channels for this question. With the transformation of blanks to zeroes, Power BI calculates schools' responses in this example as 10, 20, and 0 with an average of 10.

What is the impact?

The trend data before 2024 were adjusted to fit this methodology so the trend view is consistent. This adjustment did not change the overall findings or trend (more schools are moving towards digital marketing).