Thank you so much for posting videos on how to use the Casio fx-9750Gii for statistics! I had been struggling to use my calculator because my textbook only provides instructions for a Texas Instrument calculator. These videos are so helpful, and they are really easy to follow. I'm doing well in my statistics class because of help from these videos! I was also wondering if you had a video of how to graph formulas like "y=mx+b"? I have tried to do it on the Casio but can't seem to do it. Thank you for all your help!!
We don't currently have any videos covering plotting functions, and I don't have familiarity with this option on this calculator. Hopefully this is a topic we can cover in a future video and perhaps other math topics.
The SE shown is for the intercept, which is certainly confusing and should have been labeled more clearly in the output. The SE for the slope can be computed by taking the slope's point estimate (-1.94) and dividing it by the t-value (-7.33) to get ~0.265. Agreed that this is confusing. Doing inference on regression coefficients (or anything more complex than that) on a handheld calculator is a bit of a mess on this calculator as well as any other handheld calculators. If you're interested in trying out statistical software (the best ones are free, e.g. R or Python), you can find several labs for getting introduced to a few different softwares on our website by browsing through one or more of the resources with any of the following books: - openintro.org/book/os - openintro.org/book/ims - openintro.org/book/biostat The labs and PDFs of all of these books are also free. The ability to use statistical software such as R or Python is also an extremely useful and marketable skill. The primary reason handheld calculators are still a thing is because regular computers aren't allowed in a testing environment, and the handheld calculators still allow or relatively simple calculations.
+Steven Florindo Not seeing a great way to do this on the Casio, but here's one way: 1. Get the SE by taking the point estimate (-1.9384491) divided by the t-test statistic (-7.3321819). Basically, this is just taking the t-score formula and solving for SE: t = (estimate - null value) / SE, where the null value is 0. In the video's example, this results in SE = 0.2644. 2. Next, take the df (8) and the confidence level of interest, e.g. 95%, and get t*. You can get this through the DIST > t > Invt menu, then since I'm doing a 95% confidence interval, I set Area to 0.975 (so 0.025 is in the upper tail) and df to 8 gives a t* = 2.306. (For some reason, I'm getting a negative value when doing this, so maybe seems like the Casio here takes the upper tail, so could just use 0.025.) 3. Use the t-confidence interval formula: estimate ± t* SE -1.9384491 ± (2.306) (0.2644) (-2.55, -1.33) If you aren't tied to using a graphing calculator for class and can use a regular computer, consider learning how to use the statistical software R. While it's got a steeper learning curve, the software is open source (100% free!), and we provide free labs to get started on our website, openintro.org.
+OpenIntroOrg thank you, I'm the only one in my class with this calculator and my teacher is unfamiliar with it so these videos allow me to keep up, thanks for the help!!
Sir, id like to ask what value of alpha was that p value compared to. I just couldnt figure out how to set alpha where to compare p value results to reject or nit reject the null. Thanks in advance!
Any typical value of alpha -- from the commonly used 0.05 down to a rarely used but high bar of 0.001 -- would have led to the same conclusion to reject the null hypothesis.
@@benzfx4903 The alpha value isn't entered into the calculator for a hypothesis test: you compare the alpha value with the p-value yourself and determine whether to reject / not reject H0.
Thanks a lot, sir! I get it now. I forgot that the p value result wont be affected by the significance level. I was thinking that as we change alpha, it makes changes in p value. Changing alpha only makes change on the value we want our p value compared to.
+Mos Rus (mRussia) Sorry for the delayed reply -- our response will probably be more useful to others with the same question than you at this point. Here's a video we made on this topic: ua-cam.com/video/ISYpjr21VDA/v-deo.html
Thank you so much for posting videos on how to use the Casio fx-9750Gii for statistics! I had been struggling to use my calculator because my textbook only provides instructions for a Texas Instrument calculator. These videos are so helpful, and they are really easy to follow. I'm doing well in my statistics class because of help from these videos! I was also wondering if you had a video of how to graph formulas like "y=mx+b"? I have tried to do it on the Casio but can't seem to do it. Thank you for all your help!!
We don't currently have any videos covering plotting functions, and I don't have familiarity with this option on this calculator. Hopefully this is a topic we can cover in a future video and perhaps other math topics.
For reference, we also have a playlist for the TI-84:
ua-cam.com/play/PLkIselvEzpM7N8zVRRUl7V8aTdoTsJ919.html
Which coefficient is the SE variable for? I'm having a hell of a time writing Regression CIs with this calculator.
The SE shown is for the intercept, which is certainly confusing and should have been labeled more clearly in the output. The SE for the slope can be computed by taking the slope's point estimate (-1.94) and dividing it by the t-value (-7.33) to get ~0.265. Agreed that this is confusing.
Doing inference on regression coefficients (or anything more complex than that) on a handheld calculator is a bit of a mess on this calculator as well as any other handheld calculators. If you're interested in trying out statistical software (the best ones are free, e.g. R or Python), you can find several labs for getting introduced to a few different softwares on our website by browsing through one or more of the resources with any of the following books:
- openintro.org/book/os
- openintro.org/book/ims
- openintro.org/book/biostat
The labs and PDFs of all of these books are also free. The ability to use statistical software such as R or Python is also an extremely useful and marketable skill. The primary reason handheld calculators are still a thing is because regular computers aren't allowed in a testing environment, and the handheld calculators still allow or relatively simple calculations.
how could we construct confidence intervals for the slop of the regression line? the INTER tab only has options for 1-sample and 2-sample
+Steven Florindo
Not seeing a great way to do this on the Casio, but here's one way:
1. Get the SE by taking the point estimate (-1.9384491) divided by the t-test statistic (-7.3321819). Basically, this is just taking the t-score formula and solving for SE: t = (estimate - null value) / SE, where the null value is 0. In the video's example, this results in SE = 0.2644.
2. Next, take the df (8) and the confidence level of interest, e.g. 95%, and get t*. You can get this through the DIST > t > Invt menu, then since I'm doing a 95% confidence interval, I set Area to 0.975 (so 0.025 is in the upper tail) and df to 8 gives a t* = 2.306. (For some reason, I'm getting a negative value when doing this, so maybe seems like the Casio here takes the upper tail, so could just use 0.025.)
3. Use the t-confidence interval formula:
estimate ± t* SE
-1.9384491 ± (2.306) (0.2644)
(-2.55, -1.33)
If you aren't tied to using a graphing calculator for class and can use a regular computer, consider learning how to use the statistical software R. While it's got a steeper learning curve, the software is open source (100% free!), and we provide free labs to get started on our website, openintro.org.
+OpenIntroOrg thank you, I'm the only one in my class with this calculator and my teacher is unfamiliar with it so these videos allow me to keep up, thanks for the help!!
THANK YOU
Sir, id like to ask what value of alpha was that p value compared to. I just couldnt figure out how to set alpha where to compare p value results to reject or nit reject the null. Thanks in advance!
Any typical value of alpha -- from the commonly used 0.05 down to a rarely used but high bar of 0.001 -- would have led to the same conclusion to reject the null hypothesis.
Thanks for your response, sir but I meant if it is possible to change the alpha value in the calculator or if it can only use alpha equals 0.05
@@benzfx4903 The alpha value isn't entered into the calculator for a hypothesis test: you compare the alpha value with the p-value yourself and determine whether to reject / not reject H0.
Thanks a lot, sir! I get it now. I forgot that the p value result wont be affected by the significance level. I was thinking that as we change alpha, it makes changes in p value. Changing alpha only makes change on the value we want our p value compared to.
How would you do this on a TI-84? I am getting totally different results.
+Mos Rus (mRussia)
Sorry for the delayed reply -- our response will probably be more useful to others with the same question than you at this point.
Here's a video we made on this topic:
ua-cam.com/video/ISYpjr21VDA/v-deo.html