Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
I'm not even sure what you're talking about. I've never said my education was a waste. That's MSchemist, but big surprise that you cannot keep up with one poster vs another. I have no idea why people complain. I was employed as an associate scientist in industry before I left grad school. My opinions on HR is based on my experience obtaining an undergrad psychology degree (one of my undergrad degrees) and I know that psych majors aren't the brightest bulbs on the tree. I did a social science and a physical science side by side as an undergrad, so I have first hand experience regarding the rigor of material and quality of students in each major.
There is no comparison between an undergrad in psychology and a PhD. But continue to tell yourself the rigor is similar. I wouldn't trust a BA in psychology either. So you have the expertise in psychology of my previous students....I should definitely defer to you.
Therefore I concede, because you have gotten A's in UG psych classes, you are obviously more informed than me...despite the fact that I spent years teaching those classes, while completing my PhD.
your question was how do you falsify the hypothesis. The link posted was the "tool" as you call it that is used to falsify or support a hypothesis. I don't have a 4-year HR degree. I do not have creds from UoP.
I have a 4 year degree from a strong land grant institution and a 5 year degree from a second.
No, pay attention to what you're reading. My question was not how to falsify a hypothesis (unlike you I actually do the work in my daily routine) and you're completely wrong if you think stats are at the heart of scientific falsification, but that's exactly what social sci people do/think without fail because what you do is largely stats, and not science. That's why you don't generally have hypotheses. That's why laws, theories, and predictions do not result from your work. Anyhow, again, what's the hypothesis and how is it falsifiable? The question relates to the specific topic. I don't need a kindergarten link on statistics.
Next, I was referring to bunny's UofP creds.You responded (poorly) to my response to her post. I know you have an undergrad and grad, and I assume both are soft.
There is no comparison between an undergrad in psychology and a PhD. But continue to tell yourself the rigor is similar. I wouldn't trust a BA in psychology either. So you have the expertise in psychology of my previous students....I should definitely defer to you.
Therefore I concede, because you have gotten A's in UG psych classes, you are obviously more informed than me...despite the fact that I spent years teaching those classes, while completing my PhD.
The typical HR worker has a 4 year degree. Is that too complicated to grasp? And as we've discussed in the past don't look for validation on your organizational psychology trip from me. I just don't have it to give. I recognize that it has its place (possibly), but the attempts to elevate beyond its value is a waste of time imo. This conversation alone with the constant focus on stats cements my position. The analogy I use is about building a house. A lot goes into building a house and the use of screw guns, hammers, etc are useful and can expedite, but hammering the nail is not the point. Building the house is the point, but don't tell that to people who manufacture hammers.
No, pay attention to what you're reading. My question was not how to falsify a hypothesis (unlike you I actually do the work in my daily routine ) and you're completely wrong if you think stats are at the heart of scientific falsification, but that's exactly what social sci people do/think without fail because what you do is largely stats, and not science. That's why you don't generally have hypotheses. That's why laws, theories, and predictions do not result from your work. Anyhow, again, what's the hypothesis and how is it falsifiable? The question relates to the specific topic. I don't need a kindergarten link on statistics.
Next, I was referring to bunny's UofP creds.You responded (poorly) to my response to her post. I know you have an undergrad and grad, and I assume both are soft.
and I don't need a kindergarten link on scientific laws and theories.
A nice article comparing "hard science" research to "soft science" research. Good reading if you have the time.
Have you read it? Do you think you're going to scare me away with 13 pages LOL! Based on what I've read thus far I'm not seeing disagreement with my position. Oh, and you did not answer the question for the 4th and 5th post in a row and we both know why. Because it's bull shyte. There is no actual hypothesis and there is no actual way to falsify it.
The typical HR worker has a 4 year degree. Is that too complicated to grasp? And as we've discussed in the past don't look for validation on your organizational psychology trip from me. I just don't have it to give. I recognize that it has its place (possibly), but the attempts to elevate beyond its value is a waste of time imo. This conversation alone with the constant focus on stats cements my position. The analogy I use is about building a house. A lot goes into building a house and the use of screw guns, hammers, etc are useful and can expedite, but hammering the nail is not the point. Building the house is the point, but don't tell that to people who manufacture hammers.
The typical HR worker is not developing and implementing assessments. The typical HR worker you are talking about is called an HR generalist. They call experts on selection and assessment for guidance. My point is you can hammer your typical HR worker all you want but in almost all companies of any significant size they have experts or hire experts for guidance. Yes there are bad consulting firms out there (I personally haven't seen anything good that has come out of Kronos) but the point I am trying to hammer home to you is the typical HR worker is not the person making hiring decisions. They are there to screen based upon min. quals and send out links to assessments. The person making the hiring decisions is the hiring manager, I don't know how many times I have to say this. The hiring manager communicates with someone like me as to what they are looking for in an applicant and we discuss ways to narrow the applicant pool in an effort to get the best candidates. When there are large applicant pools assessments are often the easiest ways to do this, especially when you can't make an argument for min. quals like many low skill jobs. Again I am not a fan of personality alone, I typically like a combination of personality, situational judgment, and background experiences. But some companies for some reason choose to use just personality. In the end they are not just guessing, there are very real statistical relationships between job performance and scores on the personality sub dimensions.
Quote:
Originally Posted by Braunwyn
Have you read it? Do you think you're going to scare me away with 13 pages LOL! Based on what I've read thus far I'm not seeing disagreement with my position. Oh, and you did not answer the question for the 4th and 5th post in a row and we both know why. Because it's bull shyte. There is no actual hypothesis and there is no actual way to falsify it.
I have answered your question.
The hypothesis is that there is a relationship between a person's innate intelligence (as measured by the Watson-Glaser) and how well they do their job (let's assume it's a sales job and we are looking at revenue per month).
Not a law, not a theory, just a hypothesized relationship, which has been replicated millions of times.
Explain to me how that is not a hypothesis.
Notice how your link says continued observation.........how would you define continued observation?
Also, was not trying to scare you away. Just an article that not many people are aware of. You don't need to be so defensive
The typical HR worker is not developing and implementing assessments. The typical HR worker you are talking about is called an HR generalist. They call experts on selection and assessment for guidance. My point is you can hammer your typical HR worker all you want but in almost all companies of any significant size they have experts or hire experts for guidance. Yes there are bad consulting firms out there (I personally haven't seen anything good that has come out of Kronos) but the point I am trying to hammer home to you is the typical HR worker is not the person making hiring decisions. They are there to screen based upon min. quals and send out links to assessments. The person making the hiring decisions is the hiring manager, I don't know how many times I have to say this. The hiring manager communicates with someone like me as to what they are looking for in an applicant and we discuss ways to narrow the applicant pool in an effort to get the best candidates. When there are large applicant pools assessments are often the easiest ways to do this, especially when you can't make an argument for min. quals like many low skill jobs. Again I am not a fan of personality alone, I typically like a combination of personality, situational judgment, and background experiences. But some companies for some reason choose to use just personality. In the end they are not just guessing, there are very real statistical relationships between job performance and scores on the personality sub dimensions.
It was a typical HR worker that I responded to and there is no science here to speak of. Kronos and its Customer service assessment is not science.
Quote:
I have answered your question.
The hypothesis is that there is a relationship between a person's innate intelligence (as measured by the Watson-Glaser) and how well they do their job (let's assume it's a sales job and we are looking at revenue per month).
Or you can just assume the OP- a stock clerk at a grocery store. The test is the customer service assessment that covers team work, friendliness, etc. Either way, how do you falsify this hypothesis? You cannot falsify it because you cannot control for extraneous variables. You cannot say Sally did poorly on the test and therefore will do poorly on her job. You cannot say that Sally did poorly on her job and therefore she will do poorly on the test. You have no idea why someone may or may not do well at a job nor can you repeat any experiment under one set of conditions.
Quote:
Not a law, not a theory, just a hypothesized relationship, which has been replicated millions of times.
It's not a scientific hypothesis. And no, none of it has been replicated a million times because you cannot replicate behavior based on what I said above.
Quote:
Explain to me how that is not a hypothesis.
See above. A scientific hypothesis is based on educated observation that is independent of the observer and repeatable experiments must be generated to test the hypothesis.
Quote:
Notice how your link says continued observation.........how would you define continued observation?
Witnessing and repeating an experiment by anyone anywhere.
Quote:
Also, was not trying to scare you away. Just an article that not many people are aware of. You don't need to be so defensive
Sorry, it's the soft science folk who are always defensive because you want your field to be something it can never be and you want it to have a level of respect it simply does not deserve. To repeat for the 10th time my distaste comes from personal experience with both students and profs in academia as well as the commentary that comes out of threads like this.
I do appreciate that you finally attempted to answer the questions and I hope you have learned the difference between a scientific hypothesis and a lay one (as your example asserts).
It was a typical HR worker that I responded to and there is no science here to speak of. Kronos and its Customer service assessment is not science.
Or you can just assume the OP- a stock clerk at a grocery store. The test is the customer service assessment that covers team work, friendliness, etc. Either way, how do you falsify this hypothesis? You cannot falsify it because you cannot control for extraneous variables. You cannot say Sally did poorly on the test and therefore will do poorly on her job. You cannot say that Sally did poorly on her job and therefore she will do poorly on the test. You have no idea why someone may or may not do well at a job nor can you repeat any experiment under one set of conditions.
It's not a scientific hypothesis. And no, none of it has been replicated a million times because you cannot replicate behavior based on what I said above.
See above. A scientific hypothesis is based on educated observation that is independent of the observer and repeatable experiments must be generated to test the hypothesis.
Witnessing and repeating an experiment by anyone anywhere.
I never argued that Kronos' personality assessment was making a hypothesis. I have no idea how they developed that assessment, so I won't pretend to defend it. Let's stay on the topic of my hypothesis. You're statement was there are no hypotheses in psychology. I provided an example of one, yet you continue to go back to the personality assessment.
Experimentation and observation are two very different things. A hypothesis is not refuted or accepted based upon one individual/observation, but upon a collection of observations or running an experiment multiple times/simulations. What you are referring to is called a case study.
My hypothesis has nothing to do with Sally. My hypothesis has to do with the population, Sally is one member of said population. There will always be extraneous variables. Again, you are confusing experimentation with observation (although both will have extraneous variables, the latter will obviously have more). However, as research and science tell us a random sample will mitigate the impact that extraneous variables have on a hypothesis.
Do physicists falsify hypotheses if one particle (Sally) violates their hypothesis? That is what is referred to as measurement error, it is part of all "sciences" hard or soft.
Start thinking about people as particles. A physicist will have a large amount of measurement error if he/she decides to study one particle. Just as a psychologist will have a large amount of measurement error if he/she decides to study one person. It is when they study hundreds of millions of particles/individuals that hypotheses begin to be supported or falsified.
Behavior may vary, you are true in that statement, but that is why we as scientists admit that there is equifinality, ie multiple ways to obtain the same outcome. The test is not designed to predict intermediate behavior(s), it is designed to predict the outcome of behavior, which in this case is how well they perform their job. For example, there are multiple ways to run a race (some start slow, some fast, some maintain a constant pace), we are not pretending to know how they start the race, we are looking at how they finish the race. That is a common assumption of open systems. Again, not attempting to predict why someone does poorly (equifinality) just whether they will or will not do poorly. Why is the next step in science, commonly referred to as theory (Which as your link so eloquently demonstrated is different than a hypothesis). If you are attempting to contend that equifinality and open systems are not viable assumptions you have a few more "sciences" to throw under the bus, including; physics, economics, business, etc.
Finally, the way a test is "validated" is by giving people the test and then measuring their job performance. So after 2,000 people do both you can say that people responding in similar patterns to Sally perform this way on the job. How can a relationship exist between two variables if no "actual" relationship exists?
I can make a similar physics case. Let's say I smash 2,000 watermelons with a hammer. I smash it in 20 different ways and each time I smash the watermelon in a different way the juices land on the wall in differing areas. While each time I smash it the same way the juices may land a little differently they typically conform to the same overall pattern. I stop after 2,000 and decide there is a relationship and that I can predict with pretty good certainty where the juices will land given my hammer smash.
There is no way that I can know that the next time I smash a watemelon a given way that the juices will land in the same spot, is there? But because of 2,000 repititions I have a pretty good idea that will be the case, or that they will land within the standard error of measurement I have defined from my 2,000 repititions.
I never argued that Kronos' personality assessment was making a hypothesis. I have no idea how they developed that assessment, so I won't pretend to defend it. Let's stay on the topic of my hypothesis. You're statement was there are no hypotheses in psychology. I provided an example of one, yet you continue to go back to the personality assessment.
That's really the point, isn't it? Personality test here, personality tests there- it doesn't really matter whether there is consensus within community or not. HR will use it regardless. The people at Krogers or any other supermarket will use whatever they find because it's popular/in trend. It's what HR is doing and you can find a hundred other examples all over the net about "buyer beware" of junk science personality tests and how they are not only not scientific, but unethical. It is one of the many reasons why your field is not respected. Not that you will pay any attention whatsoever to what I'm saying here. Second, please stop lying. I don't understand the point when all of our posts are available for anyone to read. I asked about the hypothesis as it related to the topic of this thread or HR to bunny. My feelings about psychology, while aligned, are still somewhat distinct as I view HR as the half-wit demon child of psychology.
And do not pull this crap with me about taking issue with me for staying on topic. I'm not going to entirely abandon the topic of this thread. Whether I'm noting personality tests or any other behavioral test, the point likely stands anyhow.
Quote:
Experimentation and observation are two very different things. A hypothesis is not refuted or accepted based upon one individual/observation, but upon a collection of observations or running an experiment multiple times/simulations. What you are referring to is called a case study.
As it relates to a hypothesis no they are not. You cannot have a hypothesis without a reliable, suitable experiment/experimental plan to generate the event that causes the observation. A case study in my industry not only involves a hypothesis and set of experiments that surround it, but mechanistic studies, investigation into background literature and related science, and alternative experiments. What I noted was simply a matter of hypothesis testing, not a case study. Either way, you cannot run any experiment under the same conditions for human behavior and achieve the same results that would result in a prediction for other humans.
Quote:
My hypothesis has nothing to do with Sally. My hypothesis has to do with the population, Sally is one member of said population.
Boloney. Sally did not get hired due to this personality test. Really, I don't know how you buy into your own garbage. You cannot predict for anyone in your population. You cannot say that the vast majority of Sally's that performed well on their jobs will perform well on the test because they performed well on their jobs. It's ludicrous and intellectually disabled.
Quote:
There will always be extraneous variables. Again, you are confusing experimentation with observation (although both will have extraneous variables, the latter will obviously have more). However, as research and science tell us a random sample will mitigate the impact that extraneous variables have on a hypothesis.
No, that's what your soft science stat books tell you. A random sample is pointless in many, if not most, areas of science. I'm not going to test my tumor shrinking drug on a random sample and expect anything relevant. But, this is what I've been talking about all along. Random samples matter when you're trying to establish themes in a population for statistics, but not science. Your problem is that you don't understand science well enough to discern it from statistics. This will never change for you either.
Quote:
Do physicists falsify hypotheses if one particle (Sally) violates their hypothesis? That is what is referred to as measurement error, it is part of all "sciences" hard or soft.
Again, you're way off if you think you are predicting for anyone. Sally 1 or Sally^11111. You can go on about the comparison between science and statistics (not even hard vs soft), but it does not change a thing. You don't have scientific theories, laws, hypotheses, or valid experimentation. I challenge you to go to the physics forum with your comparison for the fun of it and see what those folk have to say. It should prove interesting. Everybody, even lay people, knows what you're saying is bunk in the end. Do you think the public, scientists, et al just want to bully HR? Have you stopped to consider that the majority here may be right?
That's really the point, isn't it? Personality test here, personality tests there- it doesn't really matter whether there is consensus within community or not. HR will use it regardless. The people at Krogers or any other supermarket will use whatever they find because it's popular/in trend. It's what HR is doing and you can find a hundred other examples all over the net about "buyer beware" of junk science personality tests and how they are not only not scientific, but unethical. It is one of the many reasons why your field is not respected. Not that you will pay any attention whatsoever to what I'm saying here. Second, please stop lying. I don't understand the point when all of our posts are available for anyone to read. I asked about the hypothesis as it related to the topic of this thread or HR to bunny. My feelings about psychology, while aligned, are still somewhat distinct as I view HR as the half-wit demon child of psychology.
As I've stated multiple times, I'm not a fan of personality assessments used alone. I do not advocate them for any of my clients, so you are kind of preaching to the choir. I simply came in to try to explain the statistics behind using them and why they are not illegal, nor will they ever be illegal.
Quote:
Originally Posted by Braunwyn
Boloney. Sally did not get hired due to this personality test. Really, I don't know how you buy into your own garbage. You cannot predict for anyone in your population. You cannot say that the vast majority of Sally's that performed well on their jobs will perform well on the test because they performed well on their jobs. It's ludicrous and intellectually disabled.
The fact of the matter is that even a small relationship is better than no relationship. Sally may have been an excellent employee, she may not have been (BTW the woman's name in the article was Sandy). What other way would you recommend hiring Sally with 1,000 applicants when only 10 positions are open? You may be right that she turns out to be an excellent employee (which is measurement error, not sure how you argue that being a scientist and all), but there is a stronger probability that someone that scores higher on the assessment will be a better candidate. If you could take 10 rabbits in a race of 1,000. Would you take the rabbit that has traits consistent with someone that has a 1/20 chance of winning the race or a 1/100? I'd also like to add that I have never seen one company that hires people just from personality tests. Those are used as a filter, then a select group are invited for interviews or job simulations and that is how the actual hires are made.
The only other option is pulling out of a hat. This is the difference with today's employment environment and one 30 years ago. There are more people applying for jobs in a wider area. Thirty years ago you went to work for a Kroger's in your neighborhood and got in because you knew someone there. Now companies have to worry about not violating EEOC and adverse impact, while people 15 miles away are applying to work there as well. There is no adverse impact in a personality assessment. However, adverse impact is likely to result if I only hire people I know. Unfortunately because of an insane amount of law suits companies have to cover their as&es and can no longer hire just anyone. Now they have to pass a test that does not discriminate against protected classes.
Quote:
Originally Posted by Braunwyn
No, that's what your soft science stat books tell you. A random sample is pointless in many, if not most, areas of science. I'm not going to test my tumor shrinking drug on a random sample and expect anything relevant. But, this is what I've been talking about all along. Random samples matter when you're trying to establish themes in a population for statistics, but not science. Your problem is that you don't understand science well enough to discern it from statistics. This will never change for you either.
A tumor and a random sample is a base rate issue, you are just using straw men now. A random sample for a small base rate phenomenon like a tumor is not the same thing as a random sample in a population that all has said trait. If I take a random sample of all people with a tumor is that not appropriate to mitigate all other issues, like say....the fact that someone may have diabetes, or take another prescription, or have an anti-body that reacts adversely to said drug? Or should I just test the drug on people with the tumor and diabetes and then generalize? No need for a random sample, right. Similarly if I test an intelligence test on a given population that has intelligence (hint, that is the entire population in question) I can draw inferences about that population and the DV, just like I can draw inferences about the impact the drug has on the tumor if I take a random sample of all people that have the tumor. Extraneous variables like diabetes, anti-bodies, culture, etc. are considered moderators in medicine and in "soft" sciences.
Quote:
Originally Posted by Braunwyn
Again, you're way off if you think you are predicting for anyone. Sally 1 or Sally^11111. Ylou can go on about the comparison between science and statistics (not even hard vs soft), but it does not change a thing. You don't have scientific theories, laws, hypotheses, or valid experimentation. I challenge you to go to the physics forum with your comparison for the fun of it and see what those folk have to say. It should prove interesting. Everybody, even lay people, knows what you're saying is bunk in the end. Do you think the public, scientists, et al just want to bully HR? Have you stopped to consider that the majority here may be right?
How would you explain a relationship between two variables then?
If I take 2,000 people's resting heart rates and have them all run a marathon and find that those with lower resting heart rates finish marathons faster (let's say the relationship is 0.50). Could I make a prediction on who will finish a marathon faster if I know that Sally^12 has a resting heart rate that is 3 standard deviations lower than Sally^15?
Would my prediction be more accurate than say....flipping a coin?
Let's put it this way....who would you place 100 dollars on? Sally^12, or Sally^15.
Quote:
Originally Posted by Braunwyn
Have you stopped to consider that the majority here may be right?
That's what I love about psychology.....everyone is an expert. I'm not even sure why it exists. Why take a class in something you are an expert in...not even really sure why you get a BS in it. You were already an expert, just like everyone else is.
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.
Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
