Four scales represent this construct: attitudes towards healthy relationships (alpha = .83), an average across seven Likert items that ask how important the item in question is (e.g., trust your partner). Negative response to unhealthy relationship scenarios (alpha = .70), count of whether the respondent would "ignore it" or "give another chance" when a person they were in a relationship did one of the listed items (e.g., has an explosive temper). Positive response to unhealthy relationship scenarios (alpha = .70), a count of whether the respondent would "talk with them about it" or "end the romantic relationship" when a person they were in a relationship did one of the listed items (e.g., makes negative comments about your body). Self-efficacy to use relationship skills (alpha = .77), an average across four Likert items that ask how confident the respondent would be doing the item in question is (e.g., set physical intimacy boundaries before a date).

There are two scales that are used to represent the peer pressure resistance construct: Peer resistance self-efficacy (alpha = .81), calculated as an average across four Likert items (scores ranging from 1 = very difficult to 5 = very easy) that ask how easy or difficult it is to resist peer pressure (e.g., by not giving into peer pressure at school). Concern with fitting in (alpha = .32), calculated as an average across four Likert items (scores ranging from 1 = strongly disagree to 5 = strongly agree) that ask how much the respondent agrees with statements about fitting in (e.g., you tend to follow the crowd).

There are two scales that are used to represent the perceptions of sex in media and online construct: expectations about sex shown in media, scored as the average of two Likert items (1 = strongly disagree, 5 = strongly agree), asking the respondent if they agree with a given statement (e.g. people expect to experience sex as it is portrayed in the media). Expectations about watching sex online (alpha = .77), scored as the average of two Likert items (1 = strongly disagree, 5 = strongly agree), asking the respondent if they agree with a given statement (e.g. people expect to experience sex as it is portrayed in the media).

There is one scale used to represent the problem solving construct (alpha = .68), calculated as an average across six Likert items (scores ranging from 1 = strongly disagree to 5 = strongly agree) that ask how much the respondent agrees or disagrees with a given statement (e.g., When you get what you want, its usually because you worked hard for it).

Four scales in this construct: Responsible decision making about drugs and alcohol (alpha = .92), an average across five Likert items (scores ranging from 1 = not at all important to 5 = very important) that ask how important the items are for making decisions about drugs or alcohol (e.g., how it might affect schoolwork). Responsible decision making about sex (alpha = .93), an average across five items that ask how important the items are for making decisions about doing something sexual with a partner (e.g., how an STD might affect physical health). Responsible decision making: social media behaviors (alpha = .65), an average across five yes/no items that ask if the respondent did a behavior (e.g., sent a text that they later regretted sending). Responsible decision making: attitudes toward social media (alpha = .62), an average across five items that ask how much the respondent agrees with given statements (e.g., it is safe to share your location on social media posts).

There are two scales that are used to represent the self-awareness construct: self-awareness about strengths and weaknesses (alpha = .88), calculated as an average across eight Likert items (scores ranging from 1 = very difficult to 5 = very easy) that ask how easy or difficult it is to do the item in question (e.g., knowing what I don't do well). Self-awareness about emotions (alpha = .85), calculated as an average across five Likert items (scores ranging from 1 = very difficult to 5 = very easy) that ask how easy the item in question is (e.g., knowing when my mood affects how I treat others).

There is one scale used to represent self-esteem construct (alpha = .90), calculated as an average across six Likert items (scores ranging from 1 = strongly disagree to 5 = strongly agree) that ask how much the respondent agrees or disagrees with a given statement (e.g., you feel like you are doing everything just about right).

There are two scales that are used to represent the sexual behavior intentions construct: intentions to have sex in the next three months, a single item that prompts students to respond to a 1 (yes, definitely), to 4 (No, definitely not) Likert item that asks if the student intends to have sex in the next three months, given the opportunity. Intentions to use a condom in the next three months, a single item that prompts students to respond to a 1 (yes, definitely), to 4 (No, definitely not) Likert item that asks if the student intends to use a condom if they have sex in the next three months.

There are two scales that are used to represent the sexual behavior knowledge construct: knowledge about HIV/STIs (alpha = .73), scored as the proportion of 5 items answered correctly about HIV/STIs (e.g., you can have an STD and feel healthy). Knowledge about consent (alpha = .68), scored as the proportion of 4 items answered correctly about consent (e.g., it is important to ask for sexual consent in all relationships whether or not each person has had sex before).

Four scales represent this construct: social-awareness: empathy (alpha = .84), an average across six Likert items (scores ranging from 1 = very difficult to 5 = very easy) that ask how easy or difficult it is to do the item in question (e.g., learning from people with different opinions from mine). Social-awareness: actions (alpha = .44), an average across three Likert items that ask how much the respondent agrees with the prompt (e.g., my actions affect my classmates). Negative bystander behavior (alpha = .83), calculated as a count of the number of times the respondent indicated they would "laugh or go along with" or "not say or do anything" in response to five prompts (e.g. spreading rumors about another person). Positive bystander behavior (alpha = .82), calculated as a count of the number of times the respondent indicated they would "confront the person" or "talk with a trusted adult" in response to five prompts (e.g. bullying within your group of friends).

There are two scales that are used to represent the social support construct: social support outreach (alpha = .85), calculated as an average across four Likert items (scores ranging from 1 = very unlikely to 5 = very likely) that ask how likely the respondent would do the following (e.g., talk to a trusted adult when faced with a problem that you do not know how to fix yourself). Importance of social support and engagement (alpha = .75), calculated as an average across four Likert items (scores ranging from 1 not at all important to 5 = very important) that ask how important the item in question is (e.g., having the support of your family).

There is a single item that represents the mental health construct - a 0 (poor) to 10 (excellent) Likert response to the question "how would you rate your mental health."

There is a single item that represents the relationship satisfaction construct - a 0 (poor) to 10 (excellent) Likert response to the question "how would you rate your mental health."

There are three scales that represent the sexual behavior construct: Ever had sex is a response to a single yes/no item - Have you ever had sex. Frequency of sex is an open field response to the question "In the past 3 months, how many times have you had sex." Frequency of risky sex is an open field response to the question "in the past 3 months, how many times have you had sex without using a condom."

There is a single item that represents the STI construct - a yes/no response to a question if the responded ever had an STD or STI.

There are four scales that represent the substance use construct: Vaping frequency is an open field response to a question asking frequency of vaping use in the past 30 days. Drinking frequency is an open field response to a question asking frequency of having at least one drink of alcohol in the past 30 days. Marijuana frequency is an open field response to a question asking frequency of using marijuana in the past 30 days. Other drug frequency is an open field response to a question asking frequency of using other drugs in the past 30 days.