Information is key

In data science and statistics, adding one new factor can totally change a computation. Information, when relevant, is never neutral.

The same goes for life. We need information to make good choices. If no one warned you how bad a school is, how would you know before signing up? If you're not taught about toxic behavior, how would you recognise it in a relationship? If nobody tells you what someone says behind your back, you can't know their true opinion.

Sometimes the issue isn't missing facts but missing certainty. You might see signs that someone likes you — smiles, gestures — but do you know they mean the same to the other person? Doubt changes how we act.

Back to statistics. In the early 20th century, William Sealy Gosset worked for Guinness and had to judge beer quality from very small samples. Existing methods assumed large samples and gave misleading results. Gosset understood that limited data increases uncertainty, so he created the Student's t-distribution to handle that uncertainty. He published his work under the pseudonym Student because Guinness wouldn't let its employees publish under their real name.

The law of t-statistic tells us whether a result is reliable. The formula is:

t = beta / SE(beta)

beta is the effect we estimate — what we observe.

SE(beta), the standard error, measures how uncertain that estimate is.

Think of beta as what you see and SE(beta) as how sure you are about it.

Example: if beta = 2 and SE(beta) = 0.5, then t = 4, the result looks strong. If beta = 2 but SE(beta) = 2, then t = 1, the same result now looks weak.

The number (beta) didn't change, the certainty did.

To conclude, it's very important to ask around and stay involved in conversations. Don't limit yourself to one friend group; have connections everywhere. Life-changing decisions shouldn't be made by just hoping things will work out.