Research Standards

How Howdy uses data

Howdy begins with primary administrative and institutional sources where they are available. Each article identifies the source, period, population and unit before drawing a conclusion.

Howdy calculates changes within a consistent source and definition. Data from different systems may be compared as separate signals, but they are not merged into a single total unless the populations are demonstrably compatible.

Statistical populations

Applications, admissions, visas, study permits, enrollment, SEVIS records, OPT participation and year-end permit stocks describe different stages or populations. Counts of people and program registrations, annual flows and year-end stocks, and undergraduate and graduate populations also differ. These measures are not added or treated as interchangeable unless the methodology supports it.

Applications

People or submissions recorded by an application system; applicants, admitted students and enrolled students are distinct.

Visas and permits

Administrative issuances or holders within the source's reporting period; not campus enrollment.

Enrollment

Students or program registrations reported by institutions under the source's academic-period definition; the unit must be checked.

SEVIS records

Administrative F/M records; not necessarily unique people or new arrivals.

OPT

A post-study category included in relevant Open Doors headline totals.

Year-end stock

People holding valid status at a point in time; distinct from activity during the year.

Evidence and sources

Source choice depends on the question. Administrative and primary statistical sources, including IIE Open Doors, the U.S. Department of State, SEVIS/SEVP, IRCC and Statistics Canada, document observed counts and activity under their own definitions. They do not all measure the same population.

Institutional and research reports can provide policy context, comparative findings and interpretation. Their methods and populations still need to be checked against the claim being made.

Surveys and industry research can examine stated preferences, motivations and concerns about cost, employment or safety. We consider the sample, timing, recruitment method and representativeness where reported. Stated intentions are not treated as observed enrollment or administrative outcomes.

Howdy analysis

Howdy may calculate growth rates, compare indexed trends, reconcile definitions, segment populations or synthesize evidence across sources. Charts and comparisons created by Howdy are distinguished from the data on which they rely. Source data remain attributable to their original publishers; a Howdy visualization does not make those data proprietary or original.

Practitioner observation

Insights from more than a decade of study-abroad counseling can help identify useful research questions. When practitioner experience is used, it is labeled separately and is not presented as a finding from administrative data or a representative survey.

Editorial review

Research articles are reviewed before publication for source attribution, statistical definitions, internal consistency and the distinction between source evidence and Howdy analysis. This is editorial review, not external peer review.

Bilingual publication

English and Chinese versions present the same underlying research. Each is edited for natural language rather than translated sentence by sentence. Quantitative claims, statistical populations, charts and cited evidence should remain substantively equivalent; material corrections are reflected in both versions.

Corrections

Factual and data errors are corrected after review. Material corrections to data, interpretation, source attribution or factual content are identified when appropriate. The original publication date is retained; a visible Updated date is used for a substantive correction or revision, not for copyediting, formatting, internal links or metadata maintenance.

Updating research

Policy and data articles may be reviewed when the relevant primary source releases a substantive update. The latest available release is not added automatically when its population, period or status is not comparable with the published series.