On March 11, the Ackerley Gymnasium might as well have been a theater. The introduction of each slide felt almost like a cue to react, and emphatic murmurs ricocheted off the walls. The data, containing graphical representations of the results of the Authentic Connections survey we take every year, was supposed to hold the promise of clarity, yet student reactions told a different story: one of incredulity. Amid the assembly chatter, two truths became evident:
- Data says one thing.
- Students say another.
For many students, that disconnect was immediate. “It was honestly quite surprising, especially the alcohol data,” says Karson X. ’27. “I guess it was mostly consistent with my expectation — I knew it wouldn’t be 100%, because that’s unrealistic — but it was still kind of shocking to see some students on the other side of the spectrum.”
“When students feel like their experience is not being accurately represented in the data that we share out, it sows distrust in the survey,” says Human Development and Health Department Head Julie Lutton.
So, what exactly is going wrong here?
After gathering some student data, as well as consulting with Ms. Lutton and statistics teacher Alina Badus, it can be said with complete certainty that … we aren’t completely sure.
Authentic Connections
Although we casually refer to it as the “Authentic Connections survey,” the questionnaire is actually called the High-Achieving Schools Survey. “The purpose of the survey, from the perspective of Authentic Connections, is really to help schools understand basically what are the changes that they can make within a school community in order to improve,” says Ms. Lutton.
Ms. Lutton lists two reasons why we take this survey. For one, the data that Authentic Connections gathers can reveal insights into the mental state of the Lakeside community. For another, the survey provides Lakeside with the ability to normalize our student data with other schools (about 13,000 other schools for the fall 2025 survey!).
The creators of the Authentic Connections survey primarily focus on assessing factors that, to them, can implicate the mental health of students; I say this with a particularly heavy emphasis on them. So, while the survey pretty thoroughly tries to assess community factors like inclusion and bullying, some other factors may not be as thoroughly represented because Authentic Connections doesn’t consider them as impactful to student mental health. “So for example, the Authentic Connections survey that you all took does not actually query students at all about sexual behavior other than sexting,” says Ms. Lutton, describing one of the factors Authentic Connections is less concerned with.
Aside from the fact that the survey is almost entirely non-customizable, schools have to pay in order to take it, which means the pool of schools that we are compared to is more self-selecting than we would want it to be.
As Ms. Lutton puts it, “We’re always in conversation about what we are trying to assess, why we are trying to assess it, and if the tool that we’re currently using to assess really gets us what we want. So far, the answer has been [that] it’s imperfect; it’s better than nothing, but it is certainly imperfect.”
Social Norms Theory
If you were one of the 90 people who responded to the special issue poll (thanks, by the way!), you probably remember the section of questions asking how honest you were when responding to surveys compared to how honest you felt your peers were when answering those same surveys. The following pie charts represent the breakdown of your responses:
It seems that the majority of respondents are in the mostly to always honest range (about 96.3%), but those same respondents seem to show more hesitancy to say the same for their peers, with “sometimes honestly” to “rarely honestly” making up almost half of the responses (48.8%). Of course, given this is data reflecting the thoughts of ninety students rather than the entire school body, it can’t be extrapolated in good faith to a broader claim about how all Lakesiders gauge themselves compared to their peers. However, this data potentially indicates a behavior that Ms. Lutton herself brought up during her presentation on the Authentic Connections data: social norms theory.
Social norms theory, as Ms. Lutton explains, says we often misperceive what others are doing: “In general, we tend to inaccurately perceive the behavior of others, and in fact, we tend to overestimate the rates at which other people are engaging in ‘bad’ behavior.”
So, if a student believes that everyone else is being dishonest, that perception can actually influence their own behavior — even if students actually are telling the truth most of the time.
This could, in part, explain why students reacted to the data in the way they did. When you have a large body of students potentially overestimating the amount of “bad” or risky behavior their peers are partaking in, only for the data to reflect the opposite, it’s not surprising that a lot of people were shocked.
But to chalk this all up to a misperception wouldn’t be telling the whole story. “It feels low-key insignificant,” says Brandon H. ’27. “They’re comparing us to 20-something other schools, and they explain the ‘red zone,’ but they do it using the most vague and arbitrary terms.” Nimona N. ’27 adds, “I understand this wasn’t the purpose of the data, but it kinda came off like a sensationalization of a lot of stuff.”
Addressing the Question of Anonymity
On the poll, there was a section asking respondents to write down what they believed were the reasons why students might feel hesitant to answer honestly on surveys.
13 of 42 respondents — about 31% — mentioned anonymity as a cause:
“Just in case something happens and they check your answers. You never know what might happen.”
“No poll is ever truly anonymous”
“That it somehow is traced back to us. We know there are ways to trace it back to us. We’re not stupid.”
“Because big brother is watching”
So let’s address the elephant in the room. Exactly how “secret” is the data we enter into surveys like Authentic Connections?
The Authentic Connections survey is purportedly very secure — so much so that Lakeside can’t even access all of the student data gathered. For example, Lakeside is unable to access any of the specific question-by-question data for the anxiety, depression, and rule-breaking sections of the survey (which one could argue is another limitation). Part of this is to maintain control of Authentic Connections’ intellectual property, but it’s also to maintain student anonymity.
Of course, we are putting a lot of trust into Authentic Connections. Security breaches are always a concern, and that is a risk that Lakesiders take when we send over sensitive student information in the form of our survey responses to them.
Fear of anonymity is a valid concern, especially when students are not fully aware of how their data is being used. “I don’t fully trust if it [the survey] is really anonymous, cause we don’t really know anything about the company behind it,” says Karson. “When you’re hearing on the news about all these companies having data breaches, I’m like, ‘do we really trust them?’” Indeed, Ms. Lutton agrees that “It’s hard for students, sometimes, to trust that that data really is confidential.” She even points out how her coworkers agree that they wouldn’t trust the anonymity of the Authentic Connections survey if they were in middle or high school.
No matter how well-designed or protected the survey is, perception plays a big role in how students respond. If students believe their answers can be somehow traced back to them, that belief alone is enough to influence their honesty, regardless of the actual level of security. Part of building that trust is by being direct and clear with students about how their data is being handled, as well as implementing actual technical safeguards. “[We need] to show the students that this [survey] is really anonymous: here’s how we’re going to treat the information, here’s how we care about privacy, [and] here’s who’s able to see the data,” says Ms. Badus.
The Anatomy of an Effective Survey
The data from the poll shows that students overwhelmingly prefer multiple-choice questions when it comes to participating in surveys. Out of 71 responses, 57 students (over 80%) said that multiple-choice questions encourage more participation, while only two favored free response and 9 selected both options. A few others provided more nuanced answers, such as providing multiple-choice questions with the option to explain further.
The majority of respondents also decided that optional answers were preferable over mandatory answers (about 40%), with respondents who picked mandatory answers or both tied for second (around 20% for each). About 8% of respondents elaborated that it depended on the size of the survey, with mandatory questions better for short surveys, but optional questions better for longer surveys.
At the end of the survey, some students additionally noted that surveys should be administered by teachers or advisors during a set class period. The setting and timing of when data is collected can play a huge role in how respondents interact with the survey, for better and for worse. For example, if a survey is sent out at the end of a long school day, they may be too tired to think carefully about their answers, leading them to rush through and respond less accurately. Similarly, if a survey is sent out during a particularly stressful week (like the week before spring break), answers may skew more negatively than normal. “It can be very easy to skew the data in unintentional ways … the order in which questions are presented, the setting in which you ask, the time of day… these can all make a difference,” Ms. Badus elaborates.
Analyzing and Representing Data
The analysis which is done of the Authentic Connections data is largely done by Authentic Connections themselves, and Lakeside doesn’t even get all of the data back.
However, the real place for concern is how we think about certain data in relation to other data. It is important not to correlate too many variables and get false positives. “The danger is when you have a really big dataset … if you test enough things, just by chance, you will find some positive associations somewhere … so you also want to be really careful about what you’re going to measure and how, and not overdo it,” says Ms. Badus.
This might sound counterintuitive; wouldn’t having a larger set of variables and data clarify any underlying trends from unrelated outliers? Generally speaking, yes, but only if we are clear on what information we are trying to find in the first place. What Ms. Badus is describing is called p-hacking or data dredging, which is when you repeatedly and randomly test variables until you find something statistically “significant.” In short, the key is specificity. If you have a specific goal in mind, the fewer variables and superfluous numbers you introduce, the clearer the answer you’re looking for may be.
But numbers are just … numbers (very eloquently put, I know). At some point, the issue becomes less about calculation and more about interpretation of data.
A lot of the power of data is told in its representation to the viewer. What is the context? What are we trying to highlight as a point of interest? Does it make more sense to use a bar graph or a box plot? If done correctly, data visualization can help clarify the otherwise superfluous and ambiguous information for the viewer. However, if done poorly, it can end up causing confusion and negatively affect the perceived authenticity of the data. “It’s really hard to tell specifics without exposing a lot of people, but if there was a better way to present the data, I feel like that would be much more useful,” says Nimona.
For example, one of the most misunderstood parts of the Authentic Connections survey results has to do with the categorization of “red zones.” According to Ms. Lutton, a red zone is a threshold set by Authentic Connections which identifies the percentage of students who report very high levels of behaviors like anxiety and depression. “All I can share is the percentage of students that Authentic Connections, using their own algorithms and their own statistical analysis, have deemed to be in that red zone,” says Ms. Lutton. Unfortunately, a lot of this could be clarified with the presence of question-by-question data to illuminate what percentage of Lakesiders specifically struggle with — say, anxiety — and at what levels.
“I don’t think that we did as good of a job as we should have in explaining the concept of the red zone,” says Ms. Lutton. As a result, befuddlement ensued. “Some of my advisees were a bit confused because they couldn’t remember information from slide to slide,” remarks Ms. Badus. Brandon and Nimona agree that the red zone wasn’t very clear to the student body.
Ms. Badus suggests consistency among the slides to be a way to mitigate this confusion, so that students wouldn’t forget the red zone definition or get mixed up in the data. She also comments on how visual elements like color could’ve helped clarify what students were actually looking at: “I know I really wanted the red zone to be red on those graphs. My brain just really responded to factors like that.”
Ultimately, Ms. Badus reminds us that numbers alone don’t tell the whole story: “Just because you have numbers … doesn’t mean this is the reality or truth. It’s about the story you’re telling and being honest about the limitations of this story.” The way we present our data needs to be deliberate if we want to get the right points across.
A Lakeside Mental Health Survey..?
So it’s pretty clear that the Authentic Connections survey isn’t without its problems. But what if we decided to make our own mental health survey at Lakeside? Ms. Lutton mentions that this idea has actually been floating around for a while now.
One pro is that we can get specific numbers and data in relation to the Lakeside community, allowing us to provide more relevant information to the student body. However, the cons are that we won’t be able to get the same kind of comparative analysis that Authentic Connections does, which might help contextualize the data batter. Additionally, students might not feel comfortable with a school survey due to privacy concerns and lack of trust.
Students’ feelings on the idea of a Lakeside mental health survey are, understandibly, mixed. Karson feels like having a Lakeside mental health survey would be more trustworthy than the Authentic Connections survey: “At least I’ll know it’s confined within our school, and I won’t have to worry about random people somehow finding out about it. Even if you’re not sure if it’s completely anonymous, at least you know it is anonymous to the outside world.”
Others are less convinced of this. “Especially if students know that their data is going to be presented, I feel like there is a big incentive — even more so for a Lakeside-specific survey — to be less honest,” says Brandon, stating that a Lakeside-specific survey would “self-categorize” students, making them feel more exposed. “Instead of everyone being compared to other entire schools, it would be individuals compared to other people at school.”
Nimona adds that, if we were to make a Lakeside survey, we should still incorporate some type of comparison with our neighboring schools, both public and private. “We are friends with a lot of the people who go to the public schools around here, and socially we all run in the same circles, so I think that would give a lot more information about students at Lakeside than if we just isolated it to other private schools.”
Taken together, these perspectives reveal a larger challenge: there is no single method of data collection that can satisfy all of the desires and concerns of the student body. Even Ms. Badus claims that data collection is “probably the hardest part of statistics.”
When students were asked to pick what methods would make them feel more incentivized or free to answer more honestly, we received the following data:
About 72.6% of respondents voted for seeing the results after taking the survey (53/73)
About 49.3% of respondents voted for shorter surveys (36/73)
About 49.3% of respondents voted for more interesting questions (36/73)
About 34.3% of respondents voted for incentives (raffles, prizes, etc.) (25/73)
Note that this question offered the ability for respondents to select multiple answers, hence why the percentages do not add up to 100.
The most popular combination of the provided answers was a tie between seeing the results afterward only and seeing the results paired with more interesting questions, with 13 respondents voting for each. A close second, with nine respondent votes, contains a mix of all four options.
While this data outlines a relatively clear path to make surveys more approachable, there is no approach we can take that will meet everyone’s expectations. For example, if we decide to implement the top two desired features, the students who did not vote for them or prefer additional features may feel overlooked. Preferences vary, and what makes one student feel more engaged might leave another discouraged.
Additionally some of the most desired features (e.g., seeing the results afterward), despite already being in place, don’t instantly remediate problems. The Authentic Connections data is presented during assembly annually, yet confusion and distrust still persist.
Suffice to say, even if we try to maximize the outreach of a survey, the limitations are practically unavoidable.
Do We Even Need a Mental Health Survey?
Perhaps by even trying to make these surveys work, we might be ignoring the question of whether they are necessary or not for our goals as a school. As Ms. Badus mentions, having a set purpose and motivation for collecting data is key. So, what is our motivation? To understand what factors are affecting mental health for Lakeside students? To compare Lakeside to other schools? To figure out ways to help students deal with mental health? Additionally, we should ask: is it possible to find what we are looking for through this data?
“It would probably be more helpful to have that student-specific data, but I think getting that data is basically not going to happen,” says Brandon. Even with all possible factors taken into account, a seemingly “perfect” survey can only capture so much. As Brandon mentions, “Somebody can have really bad anxiety and not be depressed or even report it, and it wouldn’t be flagged as anything.”
Instead, he suggests we focus more on creating discussion spaces and building a school atmosphere which welcomes honest discussions about mental health: “At a school like Lakeside, there are always going to be mental health issues, so I feel like trying to assess that data feels kind of redundant or out of touch. Just having spaces open for talking about mental health, and not just having reactions to these data results which are obvious, would be good.” He specifically mentions mental health chats after his chemistry quizzes and the Technology and Mental Health lecture during the history department’s current events day as examples of these spaces.
Nimona agrees and points out a nuance that surveys, like the Authentic Connections survey, often miss: “Most people at this school have some sort of anxiety, but I feel like there’s a difference between experiencing anxiety versus having an anxiety disorder. There’s a different type of anxiety that comes with going to a rigorous school than in a school which doesn’t have the same opportunities in a social or economic sense.”
So … Now What?
It’s clear that there’s no single perfect way of collecting data, and though we can find trends and endlessly analyze the data we already have, the “ideal” survey looks different for everyone. And that’s because people are unique.
Moving forward, what we need is clarity. Clarity between students and administration. Clarity about what we are even trying to do.
If the goal is to understand the factors implicating student mental health, then maybe we need to question whether data alone can truly get us there in the first place. Although data is a valuable asset to help start conversations and guide future actions, it cannot stand in for those conversations and actions themselves. Perhaps the issue isn’t that we haven’t been able to come up with the perfect survey, but that we’re relying on surveys and statistics to do a job that they were never meant to do.
Understanding students requires more than measuring them, and while numbers are a start, how we respond to them is what actually counts.
Quotes have been edited for concision and clarity.
