students of the underrepresented sex and admitted students of the underrepresented sex.²³  Using a census of all students can avoid several issues associated with sample surveys including, but not limited to:  selection of the sampling mechanism, selection of the sample size, calculation of sampling error, and using sample estimates.  If an institution intends to administer a survey to a sample population to gauge an estimate of interests and abilities, the larger the sample, the more weight OCR will accord the estimate.

OCR evaluates whether the survey is administered in a manner designed to generate high response rates and how institutions treat responses and non-responses.

OCR looks at whether institutions provide the survey in a context that encourages high response rates, and whether institutions widely publicize the survey; give students, including those participating in club or intramural sports, advance notice of the survey; and provide students adequate time to respond.  Generally, OCR accords more weight to a survey with a higher response rate than a survey with a lower response rate, and institutions may want to distribute the survey through multiple mechanisms to increase the response rate. 

For example, for enrolled students, an institution may want to administer the survey as part of a mandatory activity, such as during course registration.  If administered as part of a mandatory activity, students also should have the option of completing the survey at a later date in order to ensure that they have adequate time to respond.  Students who indicate that they wish to complete the survey at a later time should be given the opportunity to provide their contact information to enable the institution to take steps to ensure that they complete the survey.  An institution should follow-up with those students who indicate that they wish to respond in the future. 

An institution also may choose to send an email to the entire target population that includes a link to the survey.  If an institution’s assessment process includes email, OCR considers whether the institution takes appropriate cautionary measures, such as ensuring that it has accurate email addresses and that the target population has access to email.²⁴  OCR also expects institutions to take additional steps to follow-up with those who do not respond, including sending widely publicized reminder notices. 

If institutions administer the survey through a web-based distribution system, students who indicate that they have no current interest²⁵ in athletic participation should be asked to confirm their lack of interest before they exit the system.  If response rates using the methods described

²³ For example, institutions may distribute surveys to all admitted students of the underrepresented sex with acceptance letters.

²⁴ OCR also evaluates whether the survey is administered in a manner designed to ensure the accurate identity of the respondent and to protect against multiple responses by the same individual.

²⁵ Students may have, or may be unaware of whether they will have, a future interest in athletic participation.