Friend, if something is "statistically significant" at 0 > 0.05, it means (approximately and assuming the study is otherwise sound) there's only a 5% chance the result is from random chance rather than being a real effect of the variable you're studying.
If you go down the street and get mugged by a Korean, you might be tempted to believe from this datapoint that most robbers are Korean. But there's a good chance its a fluke, most robbers are something else, and statistical significance will show your sample size of one was inadequate.
But if you examine the last 10000 reported robberies and the reported ethnicity of the perp, your statistical significance will be good, and you will be able to demonstrate conclusively which ethnic groups robbers are more likely to be assigned to.
Exactly how large a sample size you need depends on your research design and on the strength of the effect of the independent variable you are looking at.
Low Information Voter
npub149…tzmp0
2025-05-07 11:48:59
in reply to nevent1q…vef8
Author Public Key
npub1494rtg3ygq4cqawymgs0q3mcj6hucvu4kmadv03s5ey2sg32df5shtzmp0Published at
2025-05-07 11:48:59Event JSON
{
"id": "9020a0fc3997ad1ad983a290a7ac9d412a77f8662ae85dcb5095e2eeb154901b",
"pubkey": "a96a35a224402b8075c4da20f0477896afcc3395b6fad63e30a648a8222a6a69",
"created_at": 1746618539,
"kind": 1,
"tags": [
[
"e",
"ac8b8b3cd9a0a449cb8072e0838fb920498eed31bac8ce79dfab94e7091ecea0",
"",
"root"
],
[
"p",
"ee11a5dff40c19a555f41fe42b48f00e618c91225622ae37b6c2bb67b76c4e49"
],
[
"p",
"a96a35a224402b8075c4da20f0477896afcc3395b6fad63e30a648a8222a6a69"
],
[
"p",
"9c2e0a870413773cb915b9c9cd2e1cbd17e53ae2b0a86aba5e5bf0fc13c420d2"
]
],
"content": "Friend, if something is \"statistically significant\" at 0 \u003e 0.05, it means (approximately and assuming the study is otherwise sound) there's only a 5% chance the result is from random chance rather than being a real effect of the variable you're studying.\n\nIf you go down the street and get mugged by a Korean, you might be tempted to believe from this datapoint that most robbers are Korean. But there's a good chance its a fluke, most robbers are something else, and statistical significance will show your sample size of one was inadequate.\n\nBut if you examine the last 10000 reported robberies and the reported ethnicity of the perp, your statistical significance will be good, and you will be able to demonstrate conclusively which ethnic groups robbers are more likely to be assigned to.\n\nExactly how large a sample size you need depends on your research design and on the strength of the effect of the independent variable you are looking at.",
"sig": "b220cbe6f048f615193c9fa2c39e43a779929414bb64f5aa25499879b8e4e6ca23bd6297dae2cf8f508ac024af3f1d5164f939f7bdd30b09d3139be48f4620db"
}