Yeah p-hacking is a huge big problem in psychology.
And it’s true. You can use sleazy approaches to get the numbers to give you the results you want. But luckily there are specific guidelines on statistics and if you use them to do p hacking. You won’t get your paper published.
The people who review the papers know what p hacking is.
And it’s also true we mostly use stat software.
But.
Before we use the software. We learn to do it by hand.
It’s imperative because this is the only way to really learn how to do it properly with software.
People are really intimidated by statistics. And I’m not going to lie and say I found it easy. I didnt.
I had very poor math skills coming in. And I spent at least 20 hours a week outside of class , going over math basics and trying to re educate myself. Because I didn’t even have a strong high school level of math proficiency.
But I wanted to do research so bad. I told myself I was going to learn it. And excel at it. And I did. But it was a lot of frustration and tears on the way. And an incredible amount of additional study hours on a topic that put me in a bad mood. I had to bribe myself with so much ice cream to stay focused.
Also ive never heard of anyone with a PhD getting a psyD later. Are you sure about that ?
Not saying you are incorrect, just saying Ive never heard of that.
People sometimes do post-docs after PhDs. That’s just more research tho.
I’m not a clinical psychologist, I just knew a lot of people who were in a traditional clinical Psych PhD program, and PsyD was one of the path options they could take, if they wanted to become a medical psychologist/prescribe medication. Idk if it was just a short cut so they didn’t have to finish the full PhD program, but they had been accepted into a PhD program.
They still had to complete basic course work including stats, which was one of the only class where we overlapped. I don’t remember anyone ever doing stats by hand. I do remember the people who were clinical usually didn’t have to collect their own data or do any experiments for their M.S., but they also were more likely to use really high level stats analyses. I remember words like bootstrapping, and it seeming very complicated.
At one point I did use a massive data set for a class project to run forecasting or something? But post grad school, a 3-way anova is about as advanced as I’ve ever needed to get. There are people who do more advanced stat work in basic science, and use R instead of the easier programs, but it’s definitely not the norm from my experience. I guess it just depends on the size of the data set you’re analyzing and what you’re trying to do with it. For example, I was talking to somebody recently who worked in infectious disease, and she mentioned that the highest level stats she ever used was a t-test.
From my own experience (maybe it’s just a U.S. thing, idk) but there are people who do very important research without using very complicated stats, and there are people who do very important research and also use very complicated stats. There’s also often a collaborative effort where one side may be more involved in generating the data to a certain point, and the other has a very niche role where they take the research relay baton and do their complex voodoo that I don’t really understand.
There are also people in public health and other fields who don’t always do their own research, but still use giant data sets to answer very important questions. Then there seem to be some people in various fields who never design their own experiments or collect their own data, but recycle huge data sets over and over again for p-hacking using very complicated stats that I don’t understand.
I admit my experience is completely dependent on how science is conducted in the U.S., but even an 8 year PhD from a prestigious institution, plus a series of never ending post docs from other prestigious institutions, and the ability to do the most complicated stats, would not make me assume that person is necessarily smart or skilled in anything other than their own niche area of research.
The same is true for an MD or MD/PhD. They might be very smart in whatever residency they completed or whatever field they achieved their PhD, but how do they actually think about new information/approach problem solving?
I feel like we all grow up hearing way too often, that we’re all smart in our own unique way, (which is probably true), but most people tend to avoid acknowledging the fact that we’re all also very dumb in our own unique way. No single person can know everything, and no human or machine is ever completely free from error.
Empathy, respect and consideration for others, along with critical thinking, and fluid problem solving skills, are usually not things people learn in a classroom or lab. Unfortunately, those skills are also undervalued in society until the moment inevitably comes when they are desperately needed. Then, big surprise, most people don’t even know how to begin to think about ways to approach a new problem, because they never really learned to think outside of a rigid and sometimes biased box.
On the rare occasion they actually allow their ego to take a break for a moment, and try to think critically about a complicated problem, the ego will often snap back into place the moment anyone questions the off target consequences or downstream effects of their idea. This is another reason it can really pay off to have an entire team people who are uniquely smart and uniquely dumb in their own way, with varying degrees of education and life experience, all working together to solve a problem.
When you use any level of education as a tool to help you continue to think critically and solve new and challenging problems, that’s very helpful for all of society. When you allow your level of education to become your impenetrable ego, and act as a blinder/shield to even consider any information that seems like it might contradict what you already believe to be true, it becomes a danger to society.
Yeah p-hacking is a huge big problem in psychology.
And it’s true. You can use sleazy approaches to get the numbers to give you the results you want. But luckily there are specific guidelines on statistics and if you use them to do p hacking. You won’t get your paper published.
The people who review the papers know what p hacking is.
And it’s also true we mostly use stat software. But.
Before we use the software. We learn to do it by hand. It’s imperative because this is the only way to really learn how to do it properly with software.
People are really intimidated by statistics. And I’m not going to lie and say I found it easy. I didnt.
I had very poor math skills coming in. And I spent at least 20 hours a week outside of class , going over math basics and trying to re educate myself. Because I didn’t even have a strong high school level of math proficiency.
But I wanted to do research so bad. I told myself I was going to learn it. And excel at it. And I did. But it was a lot of frustration and tears on the way. And an incredible amount of additional study hours on a topic that put me in a bad mood. I had to bribe myself with so much ice cream to stay focused.
Also ive never heard of anyone with a PhD getting a psyD later. Are you sure about that ? Not saying you are incorrect, just saying Ive never heard of that.
People sometimes do post-docs after PhDs. That’s just more research tho.
I’m not a clinical psychologist, I just knew a lot of people who were in a traditional clinical Psych PhD program, and PsyD was one of the path options they could take, if they wanted to become a medical psychologist/prescribe medication. Idk if it was just a short cut so they didn’t have to finish the full PhD program, but they had been accepted into a PhD program.
They still had to complete basic course work including stats, which was one of the only class where we overlapped. I don’t remember anyone ever doing stats by hand. I do remember the people who were clinical usually didn’t have to collect their own data or do any experiments for their M.S., but they also were more likely to use really high level stats analyses. I remember words like bootstrapping, and it seeming very complicated.
At one point I did use a massive data set for a class project to run forecasting or something? But post grad school, a 3-way anova is about as advanced as I’ve ever needed to get. There are people who do more advanced stat work in basic science, and use R instead of the easier programs, but it’s definitely not the norm from my experience. I guess it just depends on the size of the data set you’re analyzing and what you’re trying to do with it. For example, I was talking to somebody recently who worked in infectious disease, and she mentioned that the highest level stats she ever used was a t-test.
From my own experience (maybe it’s just a U.S. thing, idk) but there are people who do very important research without using very complicated stats, and there are people who do very important research and also use very complicated stats. There’s also often a collaborative effort where one side may be more involved in generating the data to a certain point, and the other has a very niche role where they take the research relay baton and do their complex voodoo that I don’t really understand.
There are also people in public health and other fields who don’t always do their own research, but still use giant data sets to answer very important questions. Then there seem to be some people in various fields who never design their own experiments or collect their own data, but recycle huge data sets over and over again for p-hacking using very complicated stats that I don’t understand.
I admit my experience is completely dependent on how science is conducted in the U.S., but even an 8 year PhD from a prestigious institution, plus a series of never ending post docs from other prestigious institutions, and the ability to do the most complicated stats, would not make me assume that person is necessarily smart or skilled in anything other than their own niche area of research.
The same is true for an MD or MD/PhD. They might be very smart in whatever residency they completed or whatever field they achieved their PhD, but how do they actually think about new information/approach problem solving?
I feel like we all grow up hearing way too often, that we’re all smart in our own unique way, (which is probably true), but most people tend to avoid acknowledging the fact that we’re all also very dumb in our own unique way. No single person can know everything, and no human or machine is ever completely free from error.
Empathy, respect and consideration for others, along with critical thinking, and fluid problem solving skills, are usually not things people learn in a classroom or lab. Unfortunately, those skills are also undervalued in society until the moment inevitably comes when they are desperately needed. Then, big surprise, most people don’t even know how to begin to think about ways to approach a new problem, because they never really learned to think outside of a rigid and sometimes biased box.
On the rare occasion they actually allow their ego to take a break for a moment, and try to think critically about a complicated problem, the ego will often snap back into place the moment anyone questions the off target consequences or downstream effects of their idea. This is another reason it can really pay off to have an entire team people who are uniquely smart and uniquely dumb in their own way, with varying degrees of education and life experience, all working together to solve a problem.
When you use any level of education as a tool to help you continue to think critically and solve new and challenging problems, that’s very helpful for all of society. When you allow your level of education to become your impenetrable ego, and act as a blinder/shield to even consider any information that seems like it might contradict what you already believe to be true, it becomes a danger to society.