6 ай бұрын
7 ай бұрын
8 ай бұрын
Пікірлер
@ManognaYogaEnquiry 9 сағат бұрын
India was a lot advanced in Medicine as Ayurveda which was in practice since 2500 BC. People are now again relying more on Ayurveda which is a holistic approach. Mind Body concept existed even before BC as Ayurveda considers this very strongly. Not only Ayurveda the concept of mind body is an old concept in YOGA.
@issyzxx 5 күн бұрын
In the table at the start with the data, the 3 DVs were seemingly measured 11 times in each group, right? So when you conduct the MANOVA, why does it only say "Group"? How do you differentiate between the three groups? Does the MANOVA not give me data comparing Group 1 (and 2, and 3) with DV 1, DV 2, DV 3? Thanks!
@AlexanderSwan 5 күн бұрын
That’s not what a MANOVA does - the DVs get combined into a composite for the calculations, so it compares the groups across the composite DV.
@issyzxx 4 күн бұрын
@@AlexanderSwan oh, thank you! I'm testing the affect of 1 IV on 3 DVs with 5 conditions and a control. I thought I had to do a MANOVA in order to test significance and answer my hypothesis, but what test would you recommend instead?
@AlexanderSwan 4 күн бұрын
@ yes, you would do a MANOVA
@SERDARBALCIMD 7 күн бұрын
Hi, the analysis in the violin plots are slow. I tried to change them. The module is improved more. Check it back when you have time :)
@AlexanderSwan 7 күн бұрын
Thanks for the details! I will
@yashiro3568 7 күн бұрын
This is so funny to watch thank you. you kept me interested despite being utterly uninterested and solely watching this because I am assigned to statistical treatment against my will :)
@spghose2437 14 күн бұрын
Let's say you want to predict job performance (dependent variable) and you have: Block 1: Education level (predictor 1) Block 2: Job experience and conscientiousness (predictors 2 and 3) With hindsight, what they tell in Jamovi website is to add the first predictor (Education) in "Term1" and the second (experience) and third (conscientiousness) together in "Term2". Jamovi will then provide output showing: The model statistics (R²) for Block 1 (education alone). The change in R² (R² change) when Block 2 (experience and conscientiousness) is added. The coefficients for each predictor in each block. I think this video points on multiple regression, or else you could shed light on it regarding my comment. However thank you for your detailed description which is very helpful for students.
@AlexanderSwan 6 күн бұрын
Hi there - check out my new video on JASP’s fully realized hierarchical regression functionality and rerun this regression according to your parameters and see if that makes sense.
@ariana3331-n4r 16 күн бұрын
Statistics is a night mare for me :) Thank you for making it more easy and understandable for us :)
@patrickcelleri8934 17 күн бұрын
Sometimes on the unbearably woke end of things - expected from the beta-male, ally persona.
@AlexanderSwan 16 күн бұрын
Thanks for the view and ad support :D
@ev4red1 21 күн бұрын
Thank you for this. Is it possible to get a p value for ICC in JASP. Or do you know a way to calculate it by hand with the information provided by JASP?
@AlexanderSwan 18 күн бұрын
It’s not something JASP can do, and from my cursory Google search, there’s an R package that can assist. Otherwise, it sounds like doing an ANOVA is the way to go
@ev4red1 18 күн бұрын
@@AlexanderSwan thank you!
@산토엠디하사누자만 22 күн бұрын
You talk too much
@AlexanderSwan 21 күн бұрын
Ok? Your feedback is so helpful!
@kursatuygur9205 24 күн бұрын
Can you share the sources for CFI, TLI, and RMSEA cutoffs?
@AlexanderSwan 18 күн бұрын
These are conventions and so most sources doing a search will say about the same thing.
@sabering9459 25 күн бұрын
out of date
@AlexanderSwan 25 күн бұрын
Yeah I gotta make a new video
@danielthomson9178 29 күн бұрын
Hi Prof. Swan, I have a question regarding selecting which MANOVA test. I have 5 DVs (5 factors of personality) and then 3 IVs (culture, gender and type of sport). The research question is to see if culture (multiple levels) has an effect on personality, while gender and type of sport (both two levels) play a moderating effect. Can you point me to which test I should use on G*Power and what should go in each box? Thanks so much!
@AlexanderSwan 27 күн бұрын
Overall, you should choose the main effect or interaction that you are most interested in OR the smallest effect of all the main effects/interactions. It sounds like you want to focus on the ME of culture, but you have interactions, so what makes the most sense to you
@MatviiParinov Ай бұрын
In the beginning classical PhD joke)
@arkanareswari6597 Ай бұрын
hi, i stumbled upon this vid because apparently my data is non-linear, so i cant perform simple mediation analysis in PROCESS in SPSS. Is it correct if i use GLM mediation model for simple mediation with non-linear data? (i've deleted outlier and transformed my data with log, ln, and square but path a remains non-linear).
@AlexanderSwan Ай бұрын
No, GLM is only for linear models.
@arkanareswari6597 Ай бұрын
@@AlexanderSwan ah i see. do you have any suggestion what software i can use to analyze my data?
@AlexanderSwan Ай бұрын
@ I’m sorry, I do not except for going directly to R.
@arkanareswari6597 Ай бұрын
@@AlexanderSwan i see, many thanks!
@cellphoneszosa5005 Ай бұрын
May I ask if there is a way to increase the decimal places in the pspp output e.g. say standard deviation in descriptives since it seems to default to two decimal places only?
@AlexanderSwan 29 күн бұрын
I’m not sure; check the options to see if you can increase “significant digits”
@quyntayle Ай бұрын
youre a blessing, 4yrs+ and counting! Tysm
@HRWU Ай бұрын
Hello Alex, could you please maybe interprete the relationship between f² (R²) and the calculated sample size? I'm confused by the truth, that if I have a higher R² value, which results in a higher f² too, the total sample size i need to fullfill all the alpha-/(1-beta)-values is getting lower...? what happend, if I have a model with bias, which means, one or two of the independent variables are actually not significant, but can push the R² value higher?? In this case, i might actually need more samples to justify my model, but after the F-tests output parameter, i might only need less than 20 samples... I would really appreciate if you could spend 1 min time for my amateur level question. Thanks in advance! H.
@AlexanderSwan 29 күн бұрын
Yes, effect size (R-squared) and sample size have an inverse relationship. If one is large, the other one is small.
@HRWU 29 күн бұрын
@@AlexanderSwan but e.g. if I got a R² = 0,90 from my first 5 samples after linear regression, to assume it's legit and I don't need many more samples to justify my assumption might lead to a wrong model maybe...?
@HRWU 29 күн бұрын
@@AlexanderSwan Danke sehr für die Antwort!
@AlexanderSwan 29 күн бұрын
@@HRWU what do you mean by samples? This may be a language barrier issue, but sample in statistics typically refers to a group of people or subjects. So do you have 5 groups of people or only 5 people? If you mean the first one, how many people per group? If the second one, 5 is far too small a sample for the General Linear Model, regardless if you have reached statistical significance. That’s a bias called the law of small numbers.
@HRWU 29 күн бұрын
@@AlexanderSwan well then I maybe haven't fully understood the purpose of using G*Power to calculate the sample size. I thought i mean the 2nd case you mentioned here. Let's say we have 1000 alloyed coins and I wanna find the correlation between the Nickel content (Ni %) and the measured weight m and electrical resistance R of each coin. Apparently I don't have the time to measure all the coins. So I thought I should firstly measure only like 10 coins and try to get the R² value of a approximate Modell with Ni% = a*m + b*R + c and then based on the calculated "sample size" I have to undertake accordingly (N-10) more measurements to correct the a, b & c values to a', b' and c'. Could you please point out my misunderstanding here if I naively misunderstood the meaning of "sample size"? Thanks!
@jaja3628 Ай бұрын
hello! can i have the raw data please, thankyou!
@AlexanderSwan Ай бұрын
The data is freely available if you download the package "lsj" within Jamovi.
@RabiaHussain-rp3zu Ай бұрын
Your videos are so easy to understand. Today I heard about this software and just got your tutorial videos. Very helpful for me. Is jamovi an advanced version of SPSS?
@AlexanderSwan Ай бұрын
Thanks! No, Jamovi is an open-source app with the programming language R underneath, with the point-and-click elements of SPSS. However, SPSS is a proprietary program owned by IBM and its functionality is walled-off. This software is considered an alternative
@JosefhaAgustin Ай бұрын
Hello, thank you for the great information about JAR :) I would like to ask, how we can get the linking data? thank you
@AlexanderSwan Ай бұрын
Not sure what you mean about “linking data”?
@fathimafayha8691 Ай бұрын
Next time please explain slowly
@AlexanderSwan Ай бұрын
You can use the player controls to slow down my voice/playback. You can also play/pause the video as much as you need!
@AlexanderSwan Ай бұрын
In other words: don’t treat it like a traditional lecture!
@fathimafayha8691 Ай бұрын
@@AlexanderSwan alright. Thank you
@fathimafayha8691 Ай бұрын
@@AlexanderSwan got it
@SakethMadugula Ай бұрын
How do you move the scatterplot equation and r2 value around?
@AlexanderSwan Ай бұрын
That’s a good question - I don’t think you can
@SakethMadugula Ай бұрын
@AlexanderSwan I guess it's best to just make a textbook and type it down manually to move it around the chart. I'm trying to have my iPad as my laptop replacement as most of my heavy work is usually on the desktop at work or in my office. It's crazy how all the spreadsheet apps on iOS are handicapped in someway nor the other. Numbers honestly gets the closest to non VBA/macro spreadsheet alternative to native (windows based) Excel. Thankyou for replying tho! And Happy Thanksgiving!
@mhoraigmacfarlane4868 Ай бұрын
would you be able to do a video where you have the results data set with an unknown design and work backwards to decipher what sort of test was run then create an apa write up accordingly? many of my statistic assessments have followed this structure, it would be helpful (and appreciated) to see how you would approach such a task
@AlexanderSwan Ай бұрын
That’s an interesting suggestion. I’m not sure I follow it fully though. You’re welcome to DM me on Discord or email me more info
@estegy Ай бұрын
Thank you for the video, however I think you explain to who understand already.
@AlexanderSwan Ай бұрын
It’s a tutorial for how to do something in a computer program, not a statistics course. Some bits of course demand some explanation, and yes, those sorts of concepts and nomenclature require some background knowledge, or why else would you be using the stats program?
@mhoraigmacfarlane4868 Ай бұрын
You have an upbeat and engaging delivery of statistics! Such a breath of fresh air! Sincerely, thank-you :)
@gelatalavera98 Ай бұрын
Can I use nominal variable with dummy code as independent variable?
@AlexanderSwan Ай бұрын
Yep!
@lyubomirpadalski4645 Ай бұрын
How do you know if it is significant or not? I have a value of .95
@AlexanderSwan Ай бұрын
What’s .95? Statistical significance is usually determined by comparing your p-value to your alpha level.
@Nikom33 Ай бұрын
Is there a similar video but for a moderation analysis? Thank you.
@AlexanderSwan Ай бұрын
It’s extremely similar. Rather than saying this one variable mediates these two other variables, you say this one variable moderates, or interacts with, these two other variables. You insert your stats at approximately the same places. You interpret the same as an ANOVA interaction
@Nikom33 Ай бұрын
@@AlexanderSwan thank you very much for the response.
@sandervrum3570 Ай бұрын
saved my ass man, what a great vid!
@PanLamda Ай бұрын
Thats multiple regression mate, not multivariate
@AlexanderSwan Ай бұрын
Multivariate means “multiple variables” and terms are interchangeable in stats quite a bit. The title of the video and the thumbnail capture both. But you do you
@gabrielcalderon3 2 ай бұрын
Dr. Swan, I'm an American grad-student in Europe ( struggling in statisitics at the moment) and this series of lectures along with the one you did 4 years ago have been a GODSEND. If you ever look at yourself in the mirror and wonder if it's worth to sacrifice so much of your precious time to do these videos, just close your eyes and remember all of us who are able to fulfill our rites of passage in life thanks to your tutelage. We can't thank you enough..
@AlexanderSwan 2 ай бұрын
Thanks so much for your kind words! I’m glad I could help :)
@SnehaJindal1 2 ай бұрын
MedMod does not directly yield B values, but in APA7 we are required to report them. Any suggestions on how to proceed?
@AlexanderSwan 2 ай бұрын
B values are under the “Estimate” column
@randomdudr 2 ай бұрын
I certainly see a lot of potential in this. I have an idea for you. Why not try to use this to predict forex prices and compare real data with it. I think it is a kind of controversial topic right now. FBprophet seems to have potential to become a leading trade indicator. Also, there is very good potential in medicine, supply chain management, business analysis to name a few. Thank you for this. I have started testing it out in some ways
@irisleung2300 2 ай бұрын
Halo Alex, the video is really good. I would like to ask a question about plotting a moderation graph for SVO. How can I visualize the moderator effect on the independent variable (IV), mediator, and dependent variable (DV) with lines representing low, medium, and high levels?
@AlexanderSwan 2 ай бұрын
It sounds like you want to plot the average, and then +1 SD and then -1 SD. Those are the three lines of your y variable and then you can either have a separate plot for your moderator or put all the lines on the same plot with a legend.
@guilhermeviolin 2 ай бұрын
Hello Dr. Alex. Thank you very much for your tutorial videos, they're helping me a lot in my Ph.D. I do have a question, though. Is it possible to perform a logistic regression using repeated measures from the sampled subjects? Is there a way to account for the random effects of each individual?
@AlexanderSwan 2 ай бұрын
Yes, I believe you can, but I’m not fully versed in the nuance of logistic regression. I think you want to explore variations of logistic regression in multilevel modeling.
@guilhermeviolin 2 ай бұрын
@@AlexanderSwan Thank you very much for the reply. I`ll look into it, then
@jelenaivic5448 2 ай бұрын
Saved me while writing my thesis. Thank you so much for this!
@thenisonnarciso2066 2 ай бұрын
Thank you for keeping updating the content of JASP.❤
@nafisahabdulrazaq3853 2 ай бұрын
Thank youuu ❤️ Paying 18k for uni and I have to come on KZbin for a video, this is so helpful x
@missmeshapstarrynight6960 2 ай бұрын
Omg i just wanna say tysm for this! I had a hard time grasping this lesson and finding this is sort of like a blessing 🎉
@ronalddugan7199 2 ай бұрын
Thanks!
@AlexanderSwan 2 ай бұрын
Thank you so much!
@schwalleyf 2 ай бұрын
Thank you for your great videos. One of my friend is a professor and uses JASP to teach. I prefer Jamovi and one of the main points is that Jamovi works much better with R to get a more full experience with the software. Thanks, Fraser Schwalley
@terrimiller5080 2 ай бұрын
I have two waves of data, time 1 measure and time 2 measure. I not only want to determine if there is a mean difference between the two, but are there covariates such as age, BMI, presence of cardiac disease etc. which influence the results? I was told I could do a t-test, but the RM ANOVA would be able to account for the covariates. Do the covariates just reduce the noise, or can they explain the difference?
@AlexanderSwan 2 ай бұрын
You should definitely do an RM ANOVA because those covariates have variance to be accounted for in the model. They may impact the main comparison and so you need to know that.
@psicologiajoseh 2 ай бұрын
You made the world a better place! Thanks!
@JemimahNyambe 2 ай бұрын
Next time, just get to the point
@AlexanderSwan 2 ай бұрын
Next time, watch a different video
@JemimahNyambe 2 ай бұрын
@@AlexanderSwan 😂okay
@Lettilany 2 ай бұрын
This is life-saving. Thank you SO, SO MUCH!
@夏天-h4o 3 ай бұрын
Hello Alexander，I am really confused by the small CFI and TFL in my statistics, is there any solutions to improve those two numbers let them become lager than 0.9
@AlexanderSwan 2 ай бұрын
Those small numbers are showing issues with model fit. The model is strong, but I can’t really say much more than that without looking at the model stats. Only way to fix a model is to get better data.
@graceb9240 3 ай бұрын
Thank you so much for the video! Could you please let me know how to do a moderate analysis with two moderator variables in jamovi? I know how to do it with one moderate variable though...thank you for your help in advance!
@AlexanderSwan 2 ай бұрын
You’ll want to download the medmod package in Jamovi to help you with more complex moderation analyses
@EkaitzEsteban 3 ай бұрын
Hi Alexander, thank you for featuring my FoodieStats module in your video! This module is still a work in progress, and the goal is to cover different types of analysis related to food data. I really appreciate your exploration and feedback. Looking forward to improving and expanding the module based on insights like yours!
@AlexanderSwan 3 ай бұрын
Awesome, great to hear from you! I hope my exploration leads to great things in the module. Best of luck with its expansion :)
@user-fo1cg7do8m 3 ай бұрын
The voice chnager doesn't work on Discord. I followed everything in this tutorial.
@AlexanderSwan 3 ай бұрын
Sorry to hear that - this video is def old and probably doesn’t reflect changes in the app since the initial beta release. Reach out to the devs on their discord for help!