I liked Dr. Grande's explanation of the concepts of statistical significance, effect size, and error. For example, the way he discussed inferential statistics being created due to inferring to a larger population based upon data from a sub-sample of that population; and therefore the importance of calculating the data with respect to the possibility of a 5% margin of error or "alpha". I also did not realize the importance of effect size vs. significance, and am glad that this is made more clear in this video!
I never realized how much information was covered in a statistics class until I started to watch this video. I really appreciate the time that you put in to making this video it is a great guide. I remembered a lot of things that I forgot including type I error and type II error.
I think this video in particular is very helpful because of how often people confuse the terms statistical significance and effect size. I think of the statistical significance as the "research" aspect of things, whereas the effect size shows how practical or applicable these results are. It is very possible to find that something you have tested is statistically significant, but in combination with everything else that contributes to a specific phenomenon, it has a very small effect size.
Prior to viewing the presentation, I was not educated on statistical significance and effect size. However, Dr. Grande provided an easy to follow definition and discussion on the topic. It was interesting to learn of the differences between statistical significance and effect size, in addition to the Type 1 and Type 2 errors. With more practice and further discussion I hope to strengthen my knowledge and abilities in this area.
Attaining the definition of statistical significance really helped to clarify exactly what it means and why it is important to attain the result. It is my understanding now that statistical significance is related to the result and whether or not the result is due to the treatment or random error. The probability that the random error is responsible for the results is determined. Knowing this helps us to understand why this is so important. We then need to support these numbers with a written explanation of the result and how this may impact us in the future.
I appreciate the clarification of these terms. Specifically, that the statistical significance is the probability of the result due to treatment or random error and the effect size is the measure of the amount of change in the relationship between the variables. As you point out, it is also good to remember the publishing bias toward reporting statistical significance. It warrants a closer look to determine the effect size, if present.
Thank you for the information. I understand the difference between statistical significance and effect size more clearly. It seems important to remember that correlation does not equal causation. The meaning of the null hypothesis is difficult for me to keep straight, so I appreciate that this video brought it to surface again.
Dr. Grande, thanks for providing an in depth discussion clarifying these terms. I did not know that if less than the alpha, results are considered significant.
This video refreshed statistics from undergrad, which I greatly needed. I really liked how Dr. Grande explained the error portion because we often think of error as a mistake made by the researcher or in the study, when the error is in reference to the statistics. There is a lot to learn and to understand from this video.
This was a nice refresher video of what I had learned in statistics in undergrad. It was interesting to learn that the effect size is what indicates the change in the DV. And that although there is statistical significance, it doesn't need to have a large effect size.
Before viewing this, I had forgotten what the aspect of strong, moderate, and weak relationship was. Also, I have only briefly experienced effects size, so this video was very beneficial for me to watch. Additionally, as I've stated in previous videos, errors are still something that I struggle with identifying so this video will assist me in being able to make that determination if needed. Overall, Dr. Grande did a really excellent job of explaining this information.
Dr. Grande; It was always hard for me to understand the statistical significance breakdown- you provided the information that was easy to understand. When looking at the effective size, this information is a lot (as I often get confused with Pearson's R)- your examples really help and are a great resource to go back to.
I was familiar with the majority of the information presented in this video from taking a couple classes that are similar but one thing I never really understood was the type I and II errors. This video helped me understand what they are and how to relate them to different studies. What I like about these videos is that I can go back and re listen to them as many times as I need until I finally understand the concept.
This video was very helpful in breaking down the differences between a Type 1 and Type 2 error as well as clearly explaining the difference between effect size and statistical significance.
I like that the information was broken down. It really helped me gain a better understanding on the difference between a Type 1 Error and a Type 2 Error. It was also really good that he explained what effects are significant to the DV and IV.
This video will definitely be revisited because it does provide a lot of good information. I enjoyed learning about the random error. It wasn't something that I knew about prior to watching. I now understand that statistical significance is not the same as effect size. It is good to have this information going forward because I now know to look more closely at the effect size.
Another informative video that went with this week's module. It is interesting to be remembered that a result can be statistically significant without having a large effect size.
Thank you for the video Dr. Grande. I found it very helpful that you broke down the differences between a type 1 error and a type 2 error. I always struggled with understanding which one meant what so I appreciate the explanation.
When I saw "inferential statistics" on the first slide my heart dropped from memories from undergrad. I am not strong at math or statistics. After viewing this video I did not have much more confidence in my knowledge of statistics or math. For example, I am confused on what "alpha" stands for. I have tried researching it on my own, but I still don't have a good grasp on what it means. I believe that I will need to speak to Dr. Grande and fellow students to truly have a greater understanding.
This video explained the concepts very well. The examples were helpful. I never completely grasped the Pearson's "r" before watching this video. A strong relationship is indicated by a greater than .7 result and and weak relationship is indicated by less than .3. Most of these terms are new to me but this video helped me understand ones I was familiar with.
I was thankful to watch this video because I sometimes misread the statistical significance number. I used to confuse it with a correlation. For instance, the higher the number is and the closer to 1.0 or -1.0 it is, the stronger the correlation with Pearson's r. BUT for statistical significance, it is statistically significant under 0.05 (as long as the alpha is set at 5%). I know this may sound silly, but watching this reminded me what it was again. Another way to remember is that statistical significance is the probability that it is due to random error, and not another variable or something else.
Since I have heard these terms previously but not actually worked with them, this video was helpful for me to relearn concepts such as statistical significance, effect size, and error. I wish I had a statistics course during my undergraduate coursework, but this is a good start for understanding how to evaluate research and data.
This video again just showed me how much I had forgotten from my undergraduate statistics class. I thought the explanations of the different types of errors were clear. I think this made it very easy to understand the difference between the two. I also thought that the focus on the importance of effect size was great. Sometimes we forget that there are other factors than just statistical significance.
Mandy Moore After I watched this video, I realized too how much I had forgotten from my undergrad class about stats. I agree with your comment on effect size, because as I was watching the video I was thinking the same thing.
I has been a long time since my undergrad statistics class and I remembered a lot of what this video talked about, but only as it came up. I am glad to have clarification for the concepts in this video, like the difference between statistical significance and effect size.
Michael Smith I agree! I think this video was an eye opener to just how much I had forgotten from my undergrad class! I thought he did an awesome job explaining the difference between statistical significance and effect size! They are not the same thing!
Having the errors and hypothesis explained again was very helpful for me. The null hypothesis, errors and such always catch me up, and confuse me. Again another very helpful video.
Danielle Duboski For me this video made me feel lost. I do think that it is helpful that the scientific community agrees to use .05 for most research. There will always be some error when you are using a probability sampling to represent an entire population.
Understanding statistical significance and the type of error is one of the areas in research I have more difficulties understanding and might have to watch this video again to better understand. I think more examples on type of errors would be helpful.
I am understanding statistical significance vs. effect size more by watching the video but I need a little more clarity. It would be nice to go over this more in class just so I can make sure I'm truly getting it. Dr. Grande did differentiate the two for me and now I see that they are not one in the same. Effect size seems to be of high importance. I also learned a lot about the differences in errors. That was interesting to watch.
Figuring out the statistical significant of treatment or error is going to be very important. Learning the correlation between measurements and effect size including the alpha risk was explained very well. It is good to know type 1 and 2 errors are not human but stat errors.
I definitely learned that unless a researcher's sample size is everyone in the category (every person between 18 and 24, every victim of racism, every plumber) then there will be an error. I believe that sample equations, graphs, and diagrams of Statistical Significance, Effect Size, and Error would paint a nice picture to help with understanding.
These were really difficult for me in my statistics class when I was in my undergrad. It took a lot of work for me to understand the type 1 and type 2 errors and null hypothesis and things of that nature so it was nice to hear these things explained to me again and honestly, I will probably watch this video again!
I had to watch this video more than once to get a good grasp of it. The thing that stuck out to me the most was how codependent the statistical significance and effect size are on each other.
Susan Wilder I also thought it was nice that the difference was pointed out because when the term statistical significance first came up on the video I took the term at face value not realizing that it meant something else or that effect size was more significant.
Susan Wilder I re-watched this video many times too. However, I still have a difficult time understanding the concepts. I was never strong at math so I will have to work harder to understand these concepts.
I thought this video was great. This was a great refresher on the concepts of significance, effect size, and type I an type II errors. I really liked the discussion on Error and the difference between type I and type II, especially how changing the alpha value can increase risk of a specific type of error.
I like that you clarified that statistical significance is not hard to achieve and that the real importance is the effect size. I think in research, we hope that our work does have an effect, giving a meaningful result, even if it is a small one. In regards to error, I was not knowledgeable before this video that in terms of inferential statistics there will always be an error. I assumed that when we pick our population and sizes, for example choosing to have 20 participants, that would count as 100% of our population in which we are gathering data from.
Candace Fernandez I realized that error is basically going to happen no matter how much research you do. I would like to become more familiar with inferential statistics and why error occurs.
***** I always learned in class that there was usually error, but keeping it minimal and making sure that the mistakes you see are corrected and controlled are important.
This viedo was a great refresher of inferential statistics which I had in undergrad. I did not uderstand the need for the course then and though statistic is not my favorite subject it is vitally important to understand when conducting and reading research studies. I especially like the refresher of error and the null hypothesis as this was a sturggle for me to grasp. I also now see the importance of effect size as it indicates "how much change in the DV can be explained by IV".
Regina Ames I agree in that I never cared for statistics or understood the importance! Although it can be confusing at times, it is certainly necessary in conducting research and in analyzing the literature. So, I also thought this video served as a great refresher in terms of error, statistical significance, and effect size, and found it helpful in further understanding these concepts.
I share your sentiments Regina. In undergrad, I struggled through statistics. Although I eventually grasped the concept I seemed to have forgotten some of the key concepts. Statistics is essential in conducting research, collecting data, and analyzing data. The effect size gives an understanding of change in variables, good luck with gaining an understanding.
Regina Ames I, too, share your sentiments in regards to undergrad statistics classes. Until now, I did not really see the point. I now accept that knowing about statistics is the only way to really stay informed and be able to really understand scholarly research.
This video simplified the goal of statistical significance in saying that it is used to determine whether the result is due to treatment or error. This streamlined by understanding of the concept. I found it helpful that you phrased the concept of alpha as being willing to accept being wrong 5% of the time. I liked that you outlined the Pearson's r coefficient and what numerical value indicates how strong of a relationship. I am still a little confused about Type I versus Type II errors, I will have to ask about these in class!
Statistical significance, Effect Size and Error are all-important when conducting research. One cannot be afraid to have error within their research. This is inevitable that at least 1% will be of error. It is not so much the researchers fault but it is caused by using inferential statistics. There is a Type I and Type II error. Type I being the alpha error and Type II being the beta error. This was very interesting to learn and I would like to understand more how this is applied.
Toniette Morda I like that you mentioned "one cannot be afraid to have error within their research" I think this is such an important concept to remember because errors will occur, not everything comes out perfectly.
Toniette Morda I agree one cannot be afraid of error, especially since there is no way in getting around it. Often, we are so focused on getting things "perfect" and getting it "right" and not accepting of "things just happening". I was also interested in the error portion of this video and learning that at least 1% of all research has some sort of error.
I remember taking inferential statistics when I was in my junior year of undergrad and I struggled with these concepts. I remember about null hypothesis and Type 1 and Type 2 errors however, the concept of statistical significance and effect size are fairly new to me. I also remember that there is a fair amount of mathematical equations and graphing of parabolas that goes with type 1 and type 2 errors although I do not remember a lot about them. I defiantly believe that I will need to revisit these concepts in more depth to get a firm understanding of how they can be applied to research designs.
I like that you started with the question is the result due to treatment or random error. That sums up the statistical significance part. Statistical significance only tells us what is likely to happen. There is no absolute and I need to be aware of the percent of error.
Bethany Elstrom Yes there will always be error and I needed that stated in this video. The Type I error or Alpha error and Type II or Beta Error was new for me. Explaining the correlation versus cause in more detail would be helpful.
I appreciate the distinction that was made between statistical significance and effect size, as I tend to confuse some of these concepts. This distinction also shows the importance of presenting the effect size in order to gauge the meaning of results. I had also forgotten the concept of Pearson's r, so this video was helpful in reminding me that it measures the degree of linear relationship between two variables. However, I would like to learn more about the difference between r and r 2.
Jacklin Skibicki That's the same for me as well; the difference between r and r squared was not as clear to me. I'd particularly like to know more about r squared. A good refresher overall, though, on significance and effect size.
Jacklin Skibicki I also agree that it is important to know the effect size in order to gauge the meaning of results. This was particularly interesting to learn the difference. Sometimes I read articles that have a small effect size but have significant results and this video explained that the effect size really doesn't matter to have statistical significance.
In research, we must remember that error is not always a bad thing--and that it can actually help fine-tune an experiment and is significant in its own right.
I'll be honest, I will definitely need to visit this video again! I understand and grasp the concept of statistical significance and effect size and how a result can be statistically significant, but not have a large effect size. However, when Dr. Grande started getting into Type I and Type II error I felt lost! I understand the basic concepts of each, but I believe I am thinking too hard about how you go about calculating them.
Inferential statistics Statistics has always been a difficult concept to grasp. However, this video provided the framework in understanding statistical significance. For statistics to be significant an effect size must be present. The effect size is paramount in noting the change in variables. A new piece of information is "Pearsons r" which indicate correlations in x and y. I will have to review this material again to continually refresh my knowledge of the use of variables, correlations, and inferential statistics.
***** Ashley, I think you give a great summary of the material on statistical significant and effect size. These are definitely hard concepts to grasp but reading others take-aways is helpful.
That effect size is pretty important stuff. It would be very important to know that there is a relationship between the dependent and independent variables. However, I'm glad Dr. Grande threw in there that just because there's a relationship, it does not imply causality. I would have assumed that that would mean there was a causal relationship. That's crazy that they can figure these things out.
Jeff Pincin Yes, due to effect size's importance, no matter how small, it must always be reported. This draws a more accurate picture of the research since statistical significance is easier to achieve than effect size.
Statistical significance in the one thing I remember form my research class. Just because there was a result shown from a study, does not mean that it was "significant" to the results. Significant, in research, does not mean "important" like mentioned in the video. This changed the way I even think about that word on a day-to-day basis. I don't know if this video would make sense to someone with no background in statistics. I like to have visual and verbal examples to understand things that are complicated like this. My understanding of effect size is basically that it is a measurement of relationship. How much of the change in the dependent variable can be explained by the thing you hoped would change it? Pearson's r shows that relationship as well. I try to understand these things in basic terms, but these ones were tricky! In basic terms, type I error is that the researcher presented that there was a difference, but there really wasn't. It's not that the researcher is lying, its just that because the whole population wasn't part of the study so that the results leave room for error. Type II error is the opposite. The researcher presents that there was not a significant difference, but there really was. Does that sound right?
I liked learning how important effect size is and how it relates to statistical significance. I understand the concept of error but would like to understand how it is calculated a bit more in-depth- some parts were a bit confusing.
I thought this video helped me to better understand statistical significance and effect size, which will come in handy as writing this research paper begins.
Kimberly Dixon I also think this video is going to help us when writing the paper. I know I will be watching this video multiple times when working on my research paper!
I remember statistical significance being a touchy subject when I was in college. The p≤ 0.5 seemed almost arbitrary as the cutoff for something to be claimed statistically significant.
Allowing for error in research and having a method to measure the result for statistical significance is important. Although researchers don't want to have any error it is unlikely. However the effect size indicates how much the DV is changed.
Sherrie Tilghman Because there is so much research out there, it is extremely important for the study to measure for statistical significance. Therefore, I agree that this is important. I also liked how Dr. Grande reminded us that error will occur in research. There is always room for statistical error.
As I understand a little more clearer, to have a statistical significance there has to be at least a small effect size. Change in any research is the goal.
Lots to absorb here. I feel a little unprepared in my attempt to grasp these concepts because a lot of the terminology or foundational concepts are foreign to me. I'll definitely be re-watching this video but some of the points that I gleaned are: inferential statistics leave room for error (the only way to avoid error is to test the entire population), Pearson's r is a statistic that measures linear relationship between two variables, and there are two types of errors (Type I - ALPHA, the higher the alpha value, the higher the risk for this type of error; Type II - BETA, the lower the alpha value, the higher the risk for this type of error).
What I learned new from this video was how that a higher alpha causes a type II error. I do not remember learning that and it is probably because the way to calculate that error is complicated. I can somewhat understand why this is, but not fully. I think knowing the calculation would make it more clear. I had not see the symbols for null hypothesis and alpha in a while that I've paid much attention to so it was nice to get a refresher on those. I still have some questions about Pearson's r squared and why that is used.
Alicia Zahn It sounds like watching this video might have confused you as much as it did me. I also did not remember learning that a higher alpha contributes to increased type II error. I wanted to ask about this on class next week. With respect to Pearson's r2, I think it just shows that there is a relationship between the variables, with a stronger relationship having a value closest to -1 or 1. This does not indicate that one variable necessarily caused the other or vice versa, just that a relationship exists.
Alicia Zahn Yeah, I also need to remind myself throughout watching these videos that we're only touching the surface and we will learn much more about these and their practical applications. I'm confused as well :) But we'll get this stuff, one way or another. I looked up statistical significance and effect size online so I could see some further examples of it. That was helpful as well.
Alicia Zahn When it came to the errors, I had trouble understanding how to calculate them. I too, think that if I knew the calculation it may help to better understand the errors.
Alicia Zahn I thought that a higher alpha contributed to a type I error. Under the slide that says "Type I Error", the bullet says "Higher Alpha Increases Risk". Did I miss something or did you just go crazy with your capital "I"s. :]And I'm not sure "causes" is really right, I think it is more like it increases the possibility. Of course, I am just as confused as everyone else is about all this. I guess we will find out on Tuesday. This is definitely the video that will be the focus of our discussion.
I feel slightly mixed up on type I and type II errors but I think I may just need to re-watch the video. Also, I had forgotten about effect size versus statistical significance. I didn't remember that you can have one without the other. Something can be statistically significant but not have a large effect size. I will have to remember this throughout this course.
Alaina Van Gelder I am glad I'm not the only one having the same issues!! I need a little more clarification about the errors, and I think a simplified explanation of effect size and statistical significance would be helpful as well!
I liked Dr. Grande's explanation of the concepts of statistical significance, effect size, and error. For example, the way he discussed inferential statistics being created due to inferring to a larger population based upon data from a sub-sample of that population; and therefore the importance of calculating the data with respect to the possibility of a 5% margin of error or "alpha". I also did not realize the importance of effect size vs. significance, and am glad that this is made more clear in this video!
I never realized how much information was covered in a statistics class until I started to watch this video. I really appreciate the time that you put in to making this video it is a great guide. I remembered a lot of things that I forgot including type I error and type II error.
I think this video in particular is very helpful because of how often people confuse the terms statistical significance and effect size. I think of the statistical significance as the "research" aspect of things, whereas the effect size shows how practical or applicable these results are. It is very possible to find that something you have tested is statistically significant, but in combination with everything else that contributes to a specific phenomenon, it has a very small effect size.
Prior to viewing the presentation, I was not educated on statistical significance and effect size. However, Dr. Grande provided an easy to follow definition and discussion on the topic. It was interesting to learn of the differences between statistical significance and effect size, in addition to the Type 1 and Type 2 errors. With more practice and further discussion I hope to strengthen my knowledge and abilities in this area.
Attaining the definition of statistical significance really helped to clarify exactly what it means and why it is important to attain the result. It is my understanding now that statistical significance is related to the result and whether or not the result is due to the treatment or random error. The probability that the random error is responsible for the results is determined. Knowing this helps us to understand why this is so important. We then need to support these numbers with a written explanation of the result and how this may impact us in the future.
I appreciate the clarification of these terms. Specifically, that the statistical significance is the probability of the result due to treatment or random error and the effect size is the measure of the amount of change in the relationship between the variables. As you point out, it is also good to remember the publishing bias toward reporting statistical significance. It warrants a closer look to determine the effect size, if present.
Thank you for the information. I understand the difference between statistical significance and effect size more clearly. It seems important to remember that correlation does not equal causation. The meaning of the null hypothesis is difficult for me to keep straight, so I appreciate that this video brought it to surface again.
Dr. Grande, thanks for providing an in depth discussion clarifying these terms. I did not know that if less than the alpha, results are considered significant.
This video refreshed statistics from undergrad, which I greatly needed. I really liked how Dr. Grande explained the error portion because we often think of error as a mistake made by the researcher or in the study, when the error is in reference to the statistics. There is a lot to learn and to understand from this video.
Rachel Foster Agreed! There is a great deal to learn from this video and I too had forgotten how much I don't remember from undergrad.
This was a nice refresher video of what I had learned in statistics in undergrad. It was interesting to learn that the effect size is what indicates the change in the DV. And that although there is statistical significance, it doesn't need to have a large effect size.
Before viewing this, I had forgotten what the aspect of strong, moderate, and weak relationship was. Also, I have only briefly experienced effects size, so this video was very beneficial for me to watch. Additionally, as I've stated in previous videos, errors are still something that I struggle with identifying so this video will assist me in being able to make that determination if needed. Overall, Dr. Grande did a really excellent job of explaining this information.
Dr. Grande; It was always hard for me to understand the statistical significance breakdown- you provided the information that was easy to understand. When looking at the effective size, this information is a lot (as I often get confused with Pearson's R)- your examples really help and are a great resource to go back to.
I was familiar with the majority of the information presented in this video from taking a couple classes that are similar but one thing I never really understood was the type I and II errors. This video helped me understand what they are and how to relate them to different studies. What I like about these videos is that I can go back and re listen to them as many times as I need until I finally understand the concept.
This video was very helpful in breaking down the differences between a Type 1 and Type 2 error as well as clearly explaining the difference between effect size and statistical significance.
I like that the information was broken down. It really helped me gain a better understanding on the difference between a Type 1 Error and a Type 2 Error. It was also really good that he explained what effects are significant to the DV and IV.
This video will definitely be revisited because it does provide a lot of good information. I enjoyed learning about the random error. It wasn't something that I knew about prior to watching. I now understand that statistical significance is not the same as effect size. It is good to have this information going forward because I now know to look more closely at the effect size.
Another informative video that went with this week's module. It is interesting to be remembered that a result can be statistically significant without having a large effect size.
Thank you for the video Dr. Grande. I found it very helpful that you broke down the differences between a type 1 error and a type 2 error. I always struggled with understanding which one meant what so I appreciate the explanation.
When I saw "inferential statistics" on the first slide my heart dropped from memories from undergrad. I am not strong at math or statistics. After viewing this video I did not have much more confidence in my knowledge of statistics or math. For example, I am confused on what "alpha" stands for. I have tried researching it on my own, but I still don't have a good grasp on what it means. I believe that I will need to speak to Dr. Grande and fellow students to truly have a greater understanding.
This video explained the concepts very well. The examples were helpful. I never completely grasped the Pearson's "r" before watching this video. A strong relationship is indicated by a greater than .7 result and and weak relationship is indicated by less than .3. Most of these terms are new to me but this video helped me understand ones I was familiar with.
I was thankful to watch this video because I sometimes misread the statistical significance number. I used to confuse it with a correlation. For instance, the higher the number is and the closer to 1.0 or -1.0 it is, the stronger the correlation with Pearson's r. BUT for statistical significance, it is statistically significant under 0.05 (as long as the alpha is set at 5%). I know this may sound silly, but watching this reminded me what it was again. Another way to remember is that statistical significance is the probability that it is due to random error, and not another variable or something else.
You're videps are amazing and are helping so much during my Doctorate!
Since I have heard these terms previously but not actually worked with them, this video was helpful for me to relearn concepts such as statistical significance, effect size, and error. I wish I had a statistics course during my undergraduate coursework, but this is a good start for understanding how to evaluate research and data.
This video again just showed me how much I had forgotten from my undergraduate statistics class. I thought the explanations of the different types of errors were clear. I think this made it very easy to understand the difference between the two. I also thought that the focus on the importance of effect size was great. Sometimes we forget that there are other factors than just statistical significance.
Mandy Moore I agree! It's been forever since I've taken statistics so this was a good refresher and it certainly made the difference between errors.
Mandy Moore After I watched this video, I realized too how much I had forgotten from my undergrad class about stats. I agree with your comment on effect size, because as I was watching the video I was thinking the same thing.
Rebecca Meece i agree. i did forget a lot of this from statistics in undergrad.
I has been a long time since my undergrad statistics class and I remembered a lot of what this video talked about, but only as it came up. I am glad to have clarification for the concepts in this video, like the difference between statistical significance and effect size.
Michael Smith I agree! I think this video was an eye opener to just how much I had forgotten from my undergrad class! I thought he did an awesome job explaining the difference between statistical significance and effect size! They are not the same thing!
The description of the what Type 1 and Type 2 errors helped clear up my own confusion.
Having the errors and hypothesis explained again was very helpful for me. The null hypothesis, errors and such always catch me up, and confuse me. Again another very helpful video.
Danielle Duboski For me this video made me feel lost. I do think that it is helpful that the scientific community agrees to use .05 for most research. There will always be some error when you are using a probability sampling to represent an entire population.
I am glad to know I am not the only confused person about the null hypothesis and errors, but i do agree that the video was helpful.
Understanding statistical significance and the type of error is one of the areas in research I have more difficulties understanding and might have to watch this video again to better understand. I think more examples on type of errors would be helpful.
I am understanding statistical significance vs. effect size more by watching the video but I need a little more clarity. It would be nice to go over this more in class just so I can make sure I'm truly getting it. Dr. Grande did differentiate the two for me and now I see that they are not one in the same. Effect size seems to be of high importance. I also learned a lot about the differences in errors. That was interesting to watch.
Figuring out the statistical significant of treatment or error is going to be very important. Learning the correlation between measurements and effect size including the alpha risk was explained very well. It is good to know type 1 and 2 errors are not human but stat errors.
I agree that this will be very important to know! I just hope that I catch on to this quickly so that I will not be so confused.
I definitely learned that unless a researcher's sample size is everyone in the category (every person between 18 and 24, every victim of racism, every plumber) then there will be an error. I believe that sample equations, graphs, and diagrams of Statistical Significance, Effect Size, and Error would paint a nice picture to help with understanding.
very insightful video. i realized that significance is not as huge as i expect.
I liked how thus video reviewed the different types of errors that can occur in a research experiment and can then be translated into the writing.
These were really difficult for me in my statistics class when I was in my undergrad. It took a lot of work for me to understand the type 1 and type 2 errors and null hypothesis and things of that nature so it was nice to hear these things explained to me again and honestly, I will probably watch this video again!
I had to watch this video more than once to get a good grasp of it. The thing that stuck out to me the most was how codependent the statistical significance and effect size are on each other.
Susan Wilder I also thought it was nice that the difference was pointed out because when the term statistical significance first came up on the video I took the term at face value not realizing that it meant something else or that effect size was more significant.
Susan Wilder I re-watched this video many times too. However, I still have a difficult time understanding the concepts. I was never strong at math so I will have to work harder to understand these concepts.
I thought this video was great. This was a great refresher on the concepts of significance, effect size, and type I an type II errors. I really liked the discussion on Error and the difference between type I and type II, especially how changing the alpha value can increase risk of a specific type of error.
I like that you clarified that statistical significance is not hard to achieve and that the real importance is the effect size. I think in research, we hope that our work does have an effect, giving a meaningful result, even if it is a small one. In regards to error, I was not knowledgeable before this video that in terms of inferential statistics there will always be an error. I assumed that when we pick our population and sizes, for example choosing to have 20 participants, that would count as 100% of our population in which we are gathering data from.
Candace Fernandez I realized that error is basically going to happen no matter how much research you do. I would like to become more familiar with inferential statistics and why error occurs.
Toniette Morda I agree that in doing research we want to think that our work will have a meaningful effect on the population we are observing.
***** I always learned in class that there was usually error, but keeping it minimal and making sure that the mistakes you see are corrected and controlled are important.
I totally agree with you the bottom line is the effect size. Of course, everyone would like the work they do to have some effect.
This viedo was a great refresher of inferential statistics which I had in undergrad. I did not uderstand the need for the course then and though statistic is not my favorite subject it is vitally important to understand when conducting and reading research studies. I especially like the refresher of error and the null hypothesis as this was a sturggle for me to grasp. I also now see the importance of effect size as it indicates "how much change in the DV can be explained by IV".
Regina Ames I agree in that I never cared for statistics or understood the importance! Although it can be confusing at times, it is certainly necessary in conducting research and in analyzing the literature. So, I also thought this video served as a great refresher in terms of error, statistical significance, and effect size, and found it helpful in further understanding these concepts.
I share your sentiments Regina. In undergrad, I struggled through statistics. Although I eventually grasped the concept I seemed to have forgotten some of the key concepts. Statistics is essential in conducting research, collecting data, and analyzing data. The effect size gives an understanding of change in variables, good luck with gaining an understanding.
Regina Ames I, too, share your sentiments in regards to undergrad statistics classes. Until now, I did not really see the point. I now accept that knowing about statistics is the only way to really stay informed and be able to really understand scholarly research.
This video simplified the goal of statistical significance in saying that it is used to determine whether the result is due to treatment or error. This streamlined by understanding of the concept. I found it helpful that you phrased the concept of alpha as being willing to accept being wrong 5% of the time. I liked that you outlined the Pearson's r coefficient and what numerical value indicates how strong of a relationship. I am still a little confused about Type I versus Type II errors, I will have to ask about these in class!
Statistical significance, Effect Size and Error are all-important when conducting research. One cannot be afraid to have error within their research. This is inevitable that at least 1% will be of error. It is not so much the researchers fault but it is caused by using inferential statistics. There is a Type I and Type II error. Type I being the alpha error and Type II being the beta error. This was very interesting to learn and I would like to understand more how this is applied.
Toniette Morda I like that you mentioned "one cannot be afraid to have error within their research" I think this is such an important concept to remember because errors will occur, not everything comes out perfectly.
Toniette Morda I agree one cannot be afraid of error, especially since there is no way in getting around it. Often, we are so focused on getting things "perfect" and getting it "right" and not accepting of "things just happening". I was also interested in the error portion of this video and learning that at least 1% of all research has some sort of error.
Thanks a lot Dr. Grande, this one helps me out a lot!
I remember taking inferential statistics when I was in my junior year of undergrad and I struggled with these concepts. I remember about null hypothesis and Type 1 and Type 2 errors however, the concept of statistical significance and effect size are fairly new to me. I also remember that there is a fair amount of mathematical equations and graphing of parabolas that goes with type 1 and type 2 errors although I do not remember a lot about them. I defiantly believe that I will need to revisit these concepts in more depth to get a firm understanding of how they can be applied to research designs.
I like that you started with the question is the result due to treatment or random error. That sums up the statistical significance part. Statistical significance only tells us what is likely to happen. There is no absolute and I need to be aware of the percent of error.
Bethany Elstrom Yes there will always be error and I needed that stated in this video. The Type I error or Alpha error and Type II or Beta Error was new for me. Explaining the correlation versus cause in more detail would be helpful.
It was very well put together will be looking at this one again.
I appreciate the distinction that was made between statistical significance and effect size, as I tend to confuse some of these concepts. This distinction also shows the importance of presenting the effect size in order to gauge the meaning of results. I had also forgotten the concept of Pearson's r, so this video was helpful in reminding me that it measures the degree of linear relationship between two variables. However, I would like to learn more about the difference between r and r 2.
Jacklin Skibicki That's the same for me as well; the difference between r and r squared was not as clear to me. I'd particularly like to know more about r squared. A good refresher overall, though, on significance and effect size.
Jacklin Skibicki I also agree that it is important to know the effect size in order to gauge the meaning of results. This was particularly interesting to learn the difference. Sometimes I read articles that have a small effect size but have significant results and this video explained that the effect size really doesn't matter to have statistical significance.
You explained t his so clearly using "layman" language. Thanks so much.
In research, we must remember that error is not always a bad thing--and that it can actually help fine-tune an experiment and is significant in its own right.
I'll be honest, I will definitely need to visit this video again! I understand and grasp the concept of statistical significance and effect size and how a result can be statistically significant, but not have a large effect size. However, when Dr. Grande started getting into Type I and Type II error I felt lost! I understand the basic concepts of each, but I believe I am thinking too hard about how you go about calculating them.
Danielle Maldonado Me too Danielle! I had to watch the video more than once just to make my initial post. It's a lot to grasp.
Inferential statistics
Statistics has always been a difficult concept to grasp. However, this video provided the framework in understanding statistical significance. For statistics to be significant an effect size must be present. The effect size is paramount in noting the change in variables. A new piece of information is "Pearsons r" which indicate correlations in x and y.
I will have to review this material again to continually refresh my knowledge of the use of variables, correlations, and inferential statistics.
***** Ashley, I think you give a great summary of the material on statistical significant and effect size. These are definitely hard concepts to grasp but reading others take-aways is helpful.
Thank you for this clarification. You made it very easy and simple to understand.
You're welcome!
The Pearson's r explanation helped me in understanding the
That effect size is pretty important stuff. It would be very important to know that there is a relationship between the dependent and independent variables. However, I'm glad Dr. Grande threw in there that just because there's a relationship, it does not imply causality. I would have assumed that that would mean there was a causal relationship. That's crazy that they can figure these things out.
Jeff Pincin Yes, due to effect size's importance, no matter how small, it must always be reported. This draws a more accurate picture of the research since statistical significance is easier to achieve than effect size.
Statistical significance in the one thing I remember form my research class. Just because there was a result shown from a study, does not mean that it was "significant" to the results. Significant, in research, does not mean "important" like mentioned in the video. This changed the way I even think about that word on a day-to-day basis. I don't know if this video would make sense to someone with no background in statistics. I like to have visual and verbal examples to understand things that are complicated like this. My understanding of effect size is basically that it is a measurement of relationship. How much of the change in the dependent variable can be explained by the thing you hoped would change it? Pearson's r shows that relationship as well. I try to understand these things in basic terms, but these ones were tricky! In basic terms, type I error is that the researcher presented that there was a difference, but there really wasn't. It's not that the researcher is lying, its just that because the whole population wasn't part of the study so that the results leave room for error. Type II error is the opposite. The researcher presents that there was not a significant difference, but there really was. Does that sound right?
This was really clarifying. Thank you.
THank you for this video, very helpful! And you explain the terms very sloooowly, which is really nice.
I need to revisit this video again to get a better understanding of it.
It was nice to learn what causes the different types of errors. I am still very unsure about the statistical significance!
Statistical significance vs. effect size was interesting. I did not know the effect size was more important to achieve.
I liked learning how important effect size is and how it relates to statistical significance. I understand the concept of error but would like to understand how it is calculated a bit more in-depth- some parts were a bit confusing.
I thought this video helped me to better understand statistical significance and effect size, which will come in handy as writing this research paper begins.
Kimberly Dixon I also think this video is going to help us when writing the paper. I know I will be watching this video multiple times when working on my research paper!
I remember statistical significance being a touchy subject when I was in college. The p≤ 0.5 seemed almost arbitrary as the cutoff for something to be claimed statistically significant.
Allowing for error in research and having a method to measure the result for statistical significance is important. Although researchers don't want to have any error it is unlikely. However the effect size indicates how much the DV is changed.
Sherrie Tilghman Because there is so much research out there, it is extremely important for the study to measure for statistical significance. Therefore, I agree that this is important. I also liked how Dr. Grande reminded us that error will occur in research. There is always room for statistical error.
As I understand a little more clearer, to have a statistical significance there has to be at least a small effect size. Change in any research is the goal.
Lots to absorb here. I feel a little unprepared in my attempt to grasp these concepts because a lot of the terminology or foundational concepts are foreign to me. I'll definitely be re-watching this video but some of the points that I gleaned are: inferential statistics leave room for error (the only way to avoid error is to test the entire population), Pearson's r is a statistic that measures linear relationship between two variables, and there are two types of errors (Type I - ALPHA, the higher the alpha value, the higher the risk for this type of error; Type II - BETA, the lower the alpha value, the higher the risk for this type of error).
I think it is so interesting that we can make inferences about the population using math.
What I learned new from this video was how that a higher alpha causes a type II error. I do not remember learning that and it is probably because the way to calculate that error is complicated. I can somewhat understand why this is, but not fully. I think knowing the calculation would make it more clear. I had not see the symbols for null hypothesis and alpha in a while that I've paid much attention to so it was nice to get a refresher on those. I still have some questions about Pearson's r squared and why that is used.
Alicia Zahn It sounds like watching this video might have confused you as much as it did me. I also did not remember learning that a higher alpha contributes to increased type II error. I wanted to ask about this on class next week. With respect to Pearson's r2, I think it just shows that there is a relationship between the variables, with a stronger relationship having a value closest to -1 or 1. This does not indicate that one variable necessarily caused the other or vice versa, just that a relationship exists.
Alicia Zahn Yeah, I also need to remind myself throughout watching these videos that we're only touching the surface and we will learn much more about these and their practical applications. I'm confused as well :) But we'll get this stuff, one way or another. I looked up statistical significance and effect size online so I could see some further examples of it. That was helpful as well.
Alicia Zahn When it came to the errors, I had trouble understanding how to calculate them. I too, think that if I knew the calculation it may help to better understand the errors.
Alicia Zahn I thought that a higher alpha contributed to a type I error. Under the slide that says "Type I Error", the bullet says "Higher Alpha Increases Risk". Did I miss something or did you just go crazy with your capital "I"s. :]And I'm not sure "causes" is really right, I think it is more like it increases the possibility. Of course, I am just as confused as everyone else is about all this. I guess we will find out on Tuesday. This is definitely the video that will be the focus of our discussion.
Helpful. Thanks
I feel slightly mixed up on type I and type II errors but I think I may just need to re-watch the video. Also, I had forgotten about effect size versus statistical significance. I didn't remember that you can have one without the other. Something can be statistically significant but not have a large effect size. I will have to remember this throughout this course.
Alaina Van Gelder I am glad I'm not the only one having the same issues!! I need a little more clarification about the errors, and I think a simplified explanation of effect size and statistical significance would be helpful as well!
I thought it was interesting that you can have statistical significance and a small effect size.