In our experience surveying more than 10,000 youth participants, I-statements elicit a stronger personal connection and make more sense to youth respondents because they can easily “try the statement on” – resulting in more honest responses that better reflect youth outcomes. For example, we use “I see myself as a leader” or “I feel I have a lot to be proud of” instead of “Do you see yourself as a leader?” or “Do you feel you have a lot to be proud of?”

We strive to keep our outcome measures brief, with most consisting of approximately ten items. In cases where a program wishes to measure two outcomes (e.g. self-esteem and social competence), the resulting survey will be about 20 items long. This cuts down on time spent completing surveys, keeps youth engaged in the process, and decreases the risk of less reliable data that can result from youth scrolling through a long survey as quickly as possible without really reading and considering the items or response options. Limiting the number of outcome measures also leaves room to incorporate additional process-related questions, such as level of interest in a new initiative, satisfaction with an existing program, or the best day of the week to schedule a special event; or questions of special interest to funders.

Positively phrased survey items keep the focus on strengths and not on weaknesses, while still allowing a youth to indicate that they do not currently experience the positive attribute being referenced by disagreeing with the statement. Negatively worded items are of questionable utility in surveying youth (Jackson Barnette, 2000), can project a diagnostic sentiment, and create a survey completion experience that can elicit negative emotions or alienate youth. For example, the survey items “I feel respected at my program” or “I feel useful” encourage the respondent to reflect on positive attributes that they may already possess or be developing through the program. This is in stark contrast to the deficit-based items of “I do not feel respected at my program” and “I feel useless.”

Evaluation is not a one-person job left only to evaluators, but rather a key part of the program fabric. Front line staff are among the greatest strengths of quality sport programs. Youth-adult relationships are an important determinant of positive youth outcomes, and many youth develop meaningful and long-term relationships with their coaches and other program leaders. As such, coaches and program leaders are often better positioned to pass out tablets for program evaluation purposes or to work through simple surveys with very young participants than management-level or evaluation staff. At best, data collection is a collaborative effort involving each of the above parties – coaches, managers and evaluators – to maximize response rates and data quality, and ensure comprehension of instructions.

Current research suggests youth are more honest with technology than they are face-to-face with another human (Radovic et al., 2018). Given our increased reliance on digital technology for everything from social interaction to banking to education, this is unsurprising. We can capitalize on this reality by collecting data through engaging digital platforms that incorporate elements of gamification and competition, much like youth’s favorite mobile games and apps. A SurveyMonkey questionnaire achieves part of the aim but presents a bland and unappealing design. Using more enticing mobile-friendly platforms can result in greater enjoyment for youth respondents, increased response rates, and increased buy-in from staff when youth complete program evaluations on their own time – saving valuable program time. At MLSE LaunchPad, on average, 2/3 of youth complete their pre-program survey independently in advance of the first program session. On average, we achieve an 86% response rate.

A Likert scale is a type of rating scale used to measure attitudes or opinions. Respondents are asked to rate items based on level of agreement, frequency of a thought or action, or importance of the survey item (Iowa State University, 2010). Youth prefer Likert scales over other survey formats, finding them easier to complete (van Laerhoven et al., 2007). However, within a five-point Likert scale, youth tend towards a neutral point when it is offered as an option (Dalal et al., 2014). Based on our experiences, MLSE LaunchPad uses 4-point scales with response options ranging from Strongly Disagree to Strongly Agree, or Yes/No questions in cases where the nuanced response that a Likert scale allows is not required.