Graduate Insights makes it easy to pull a number. These practices help you make sure the number means what you think it does, especially before it goes into a report, presentation or accreditation self-study.
Before you draw conclusions
Check the sample size. Small cohorts produce noisy results. An average salary based on 12 graduates is not the same as one based on 400. Treat any figure from fewer than 30 records with caution, and note the sample size whenever you share it externally.
Choose recency or depth. Recent cohorts reflect today's labor market. Older cohorts give longer outcome windows, such as five-year rather than one-year salaries. Program review usually needs depth; enrollment marketing usually needs recency.
Work at the program level for program decisions. Institution-wide figures can hide wide variation. An 87% employment rate overall may include strong programs and programs well below the average.
Be ready to explain the methodology. If the data goes into an accreditation self-study, a state report or any external communication, you should be able to describe where it comes from. Steppingblocks data draws on more than 150 million workforce profiles. See Where does our data come from?, or ask your Partner Success Team Member for a methodology summary for your use case.
As you explore
Start broad, then narrow. Look at a college, school or program before drilling into subgroups. Patterns at the top level tell you where to focus.
Segment by graduation year, not enrollment year. Tying outcomes to when students finished gives a cleaner window for measuring what happened after graduation.
Triangulate before you conclude. One data point is a signal; two that agree start to be evidence. If a result surprises you, check it against another source or time window before building a story around it.
Document your pull as you go. Note the filters, cohort window and date of each analysis. Accreditation work in particular may require you to reproduce or explain it months later.
When you share
Share findings with context. Include the cohort definition, time window, sample size and any significant caveats. Context builds credibility and prevents misreading downstream.
Not sure where to start, or want a walkthrough of your institution's data? Your Partner Success Team Member can run a working session tailored to your goals.