web analytics
Babylon, Lectionary

Crap science is a consequence.

Matt Briggs is enjoying himself this morning, because a famous research psychologist has had to withdraw almost all his eating studies. The reason: he p-hacked. Finding non signficant results where there should be results, or considering statistics as a proof instead of a measure of improbability.

And all his false facts have to be retracted. Which is good. It also leaves many policies of the food industry in tatters.

However, any work on the unpalatable truths has to be censored, while crap science that promotes the narrative of the day is promoted.

But building a structure on ideology was the error of Lysenko. It did not work for the Soviets. It will not work now.

Crap science is a consequence of abandoning truth as surely as liberal churches die because they no longer have the Spirit, and the form of religion without the power leads to corruption, which revolts anyone who seeks truth, beauty, and the divine behind this.

Good science requires you deal with the results you don’t like. You question what you are doing. You know that you err. It finds things that are not comfortable. But it is not able to be managed or planned, and as such the political class rightly mistrust those who still have honour.

And our God of Justice notes this.

Hosea 4:1-10

Hear the word of the Lord, O children of Israel,
for the Lord has a controversy with the inhabitants of the land.
There is no faithfulness or steadfast love,
and no knowledge of God in the land;
there is swearing, lying, murder, stealing, and committing adultery;
they break all bounds, and bloodshed follows bloodshed.
Therefore the land mourns,
and all who dwell in it languish,
and also the beasts of the field
and the birds of the heavens,
and even the fish of the sea are taken away.
Yet let no one contend,
and let none accuse,
for with you is my contention, O priest.
You shall stumble by day;
the prophet also shall stumble with you by night;
and I will destroy your mother.
My people are destroyed for lack of knowledge;
because you have rejected knowledge,
I reject you from being a priest to me.
And since you have forgotten the law of your God,
I also will forget your children.
The more they increased,
the more they sinned against me;
I will change their glory into shame.
They feed on the sin of my people;
they are greedy for their iniquity.
And it shall be like people, like priest;
I will punish them for their ways
and repay them for their deeds.
They shall eat, but not be satisfied;
they shall play the whore, but not multiply,
because they have forsaken the Lord
to che

The problem with this time is that the elite have decided that the people of this nation must go. They will elect another people. But this will stop working.

Think if it as micro-evolution in action. The women who have swallowed the liberal lies: that your career is needed, that you should delay or not have children, that there is nothing in this earth but your pleasure… have zero to one children. But the women of faith and those who love children still have them.

And they will be the generation that turns up. The liberals now hate their nation for they see that the younger generation hate them, so they want a new people who will be grateful for letting them in. But they will not

Regardless of how hard they lie.

Do not lie, particularly to yourself, for your lies make you blind. They remove truth, and when there is no truth there is no joy. This is the dull, grey, hell that the narrative preaches: but that is the ur-lie of our age. The crap science that we have is a consequence of the corruption of the academy, and the management of research.

But we do not have to accept lies because p is less than 0.05 on a cherry picked outcome.

We can not be like them.


  1. Well said. I’ve been under the impression for years that there are “powers that be” in institutional, especially state sponsored, research that more or less ensure a preordained result. I started thinking this when I saw a study from Minnesota that claimed that “abstinence based sex ed doesn’t work” without even deriving a control sample, and of course performed no statistical test. They’d written the hypothesis so it was unfalsifiable if as many as one person had sex before marriage.

    When I contrived an admittedly flawed control, I found a great p value, << 0.05. Not enough to prove the point–there were other hypotheses I could point to–but enough to say "as a MN taxpayer, I'd like my money back."

  2. And of course, I concede that my comment was precisely about using p, but it was where the study was designed to avoid even that test. I’ve known for decades that certain people more or less “fish for hypotheses” by testing preexisting data this way and that until they find p < 0.05. Lost in the equation is how many hypotheses they tested before finding something, and the need for another better controlled experiment to see if it could be duplicated.

    Alas, we've found a lot of the easy stuff, and now we're in the realm of where we need the statisticians, and praise God for that. But we need to remember not to evict the logicians.

  3. pukeko Author

    I don’t need to tell Bike to use analysis of variance, and account for missing data. Personally, I think you should not do statistics unless you can do it from the command line and know precisely what you are doing.

Leave a Reply

%d bloggers like this: