Factorial Designs: Main Effects & Interactions

Рет қаралды 63,055

S. Wesley Beckwith

4 жыл бұрын

Basics of factorial designs including main effects and interactions.

Пікірлер: 43

@loxleymoon 3 ай бұрын

I wish every KZfaq video was made this clearly and concisely.

@emilyc3465 3 жыл бұрын

I'm taking research methods this semester and my professor does the least. He has no actual lectures, but will type them up and give us power points. I want to thank you for sharing this video, because this helps me out a lot.

@s.wesleybeckwith3561 3 жыл бұрын

Happy to help, and I'm sorry you are in that situation. If there are other concepts a video on would help, let me know, and if I have time I'll try to get one recorded.

@jamie6506 8 ай бұрын

Same boat. This guy needs to change jobs or retire

@indigrace8388 Жыл бұрын

I was having a lot of trouble understanding factorial designs and why we use the table format to figure out the data but you explained it so well! Thank you so much!!!

@humanperson1881 4 ай бұрын

Thank you for this. This helped me a lot to prepare for my Research exam!

@ritadangelo8808 2 жыл бұрын

The most simple description I've ever heard! Nice job!

@juice6910 3 жыл бұрын

Very good explanation of factorial designs! Great stuff man!

@xondiego Жыл бұрын

Best explanation so far about interactions I have found, Thanks for this content!!!!

@taiyabbrown 7 ай бұрын

this is so helpful in my research methods class thanks so much

@Anka1610 3 жыл бұрын

This is such a great explanation! I love how you go over all possibilities based on some tangible examples. Thank you for making this

@s.wesleybeckwith3561 3 жыл бұрын

You're very welcome!

@nikerracoleman4034 2 жыл бұрын

Thank you so much, I watched at least 100 videos and had no idea what they were saying BUT YOU MY GOOD SIR, HAVE DONE IT!! I UNDERSTAND!! I WAS CRYING BUT YOU HAVE MADE MY DAY!!!

@s.wesleybeckwith3561 2 жыл бұрын

Happy to help!

@ramagehris8836 2 жыл бұрын

Thank you! This was a super clear explanation.

@smac1130 Жыл бұрын

Thank you! You've helped me so much!!!

@Dave-lc3cd 2 жыл бұрын

this is quality content, thank you so much!

@nahashonosinde9163 4 жыл бұрын

Awesome video. Well explained!

@JulietTheLifeCoach Жыл бұрын

Great explanation, THANK YOU!

@tristanwentorth0275 2 жыл бұрын

Thank you so much for the help! I'm in a class on this and the prof is useless, really appreciate people like you!

@s.wesleybeckwith3561 2 жыл бұрын

Happy to help, and I'm sorry you are in that situation. If there are other concepts a video on would help, let me know, and if I have time I'll try to get one recorded.

@locustphysalis3300 2 жыл бұрын

Extremely helpful, thanks

@Life-xn1iq 18 күн бұрын

That you expressed in best way!

@jpeter313 9 ай бұрын

thank you so much! this was so helpful :)

@abderhmanmanchawi5277 3 жыл бұрын

Thank you. great!

@ykennedy5054 3 жыл бұрын

Thanks mate!

@s.wesleybeckwith3561 3 жыл бұрын

You're welcome!

@Evelyn-pl3we 4 жыл бұрын

Omg this was SO helpful and straightforward! Do you think you can do a video showing a 2x2x2, 2x3, and 3x3 design?

@s.wesleybeckwith3561 4 жыл бұрын

I can/will, but it might take me a bit to get to it. I'm currently swamped transitioning all of my classes to online delivery. Do you need this for a course on a tight timeline? Also, are you just looking for examples of these designs? Or what main effects and interactions might look like in these designs as well?

@Evelyn-pl3we 4 жыл бұрын

@@s.wesleybeckwith3561 I apologize for the slow response - thing's got a bit haptic due to the pandemic! Any examples would be lovely, as I am finding it challenging to grasp and work through factorial designs like 2x3x3, 3x3x3, and so on and so forth.

@raianrazal 3 жыл бұрын

Yes please more impacts of other factorial designs. I am quite new to experimental methods so this video was very helpful!

@s.wesleybeckwith3561 3 жыл бұрын

@@Evelyn-pl3we @Raian Razal I'll work on getting something recorded.

@georgeallen846 Жыл бұрын

At 4:45, why are the lines crossed? Does it matter? Can you draw them parallel or no?

@s.wesleybeckwith3561 Жыл бұрын

They are only crossed to try to show that one line is not "above" the other. The slope of those lines is not intended to be "significantly" different and the crossing can be see as some random noise.

@TheOfficeGuysTutorials 3 жыл бұрын

I did some experiments with 2 Factors (Let say A and B) and 3 Levels (A1,A2, A3 & B1, B2, B3) A1 B1 A2 B1 A3 B1 A1 B2 A2 B2 A3 B2 A1 B3 A2 B3 A3 B3 I have performed a total of 9 experiments. How can I perform a factorial analysis for this kind of experimentation.

@s.wesleybeckwith3561 3 жыл бұрын

To be able to select or recommend an analysis, I would need more information. Whether one or both factors are manipulated in a between-subjects or a within-subjects manner affects what analysis is appropriate as does the properties of your dependent variable (measurement scale, normally distributed, etc.). I’m also assuming you meant 9 different conditions, not experiments. If both factors are between-subjects and your dependent variable is normally distributed on either an interval or ratio scale of measurement, a two-way ANOVA would be an appropriate analysis and “standard” analysis in many biological and behavioral fields. If one of your IV’s is within-subjects, you would go with a mixed-factorial ANOVA. Most stats programs worth their salt (SPSS, R, SAS, MatLab) should be able to handle it. There is also a version for two repeated measures factors. However, with a 3x3 design, carry over and order effects are usually too much of a concern, and I rarely see it used. With a 3x3 design, you would also want to plan out your post-hoc tests in advance to follow up any significant effects. Depending on your a priori hypotheses, which post-hoc test you use for pair-wise comparisons and which comparisons you make vary considerably, and I’m hesitant to suggest anything without knowing more. One other major concern for post hoc tests is avoiding confounded comparisons. What this means is you shouldn’t compare any two individual conditions which have more than one difference between them. For example, comparing condition A1B1 to condition A2B2 is a confounded comparison. If there is a mean difference between these two conditions, you are unable to say it is due to A1 changing to A2 or B1 changing to B2. The comparison is confounded because there are two possible explanations for the difference you see. P.S. It looks like you have an interesting channel. I have a couple of assignments where I have students create info-graphics. I’m going to watch your videos on that to see if they would be a good resource for them.

@rababshahzad 3 жыл бұрын

I wish i could give you a hug 😭😭😭😭😢

@s.wesleybeckwith3561 3 жыл бұрын

I'm glad it helped!

@benenecat 2 жыл бұрын

A question on my homework just shows the line graphs of a crossover interaction, parallel lines (no interaction correct?), and a spreading interaction. The questions ask for each one, how many main effects there are for each 2x2 factorial design. Does every 2x2 factorial design have 2 main effects? I feel like this question is made to trick me! Thanks for your video!

@s.wesleybeckwith3561 2 жыл бұрын

A main effect specifically refers to the main effect of one IV alone. A 2x2 design has 2 IV's each with 2 levels. It is POSSIBLE to have a maximum of 2 main effects and a minimum of 0, and as I stated in the video, parallel lines indicate the absence of an interaction. I'm not going to straight up answer the homework question for you, but I do think you may be overthinking the question. One simple way/tactic to determine if their is a main effect is to ask the question, would one of the IV's all by itself change the DV scores all in same direction.

@prateekshasingh4237 10 ай бұрын

How many interaction effects for this design?

@benenecat 10 ай бұрын

@prateekshasingh4237 since asking this question I have graduated college

@prateekshasingh4237 10 ай бұрын

@@benenecat same for me 😃😂