ASA Connect

 View Only
Expand all | Collapse all

Time to end TI-84 calculator?

  • 1.  Time to end TI-84 calculator?

    Posted 07-13-2017 12:43

    The Ti84 Calculator seems to have a lock on high school AP statistics classes, and the exam.  We don't use them in our college classes.  Isn't it time that high schools use Excel more for statistics?  Excel has all the probability distributions – t-,z F, chi-square – and it far more practical.

     

    Thomas W. Ilvento

    Department of Applied Economics and Statistics

    Professor, Chair and StatLab Coordinator

    213 Townsend Hall

    Newark, DE 19716

    Office: 302-831-6773

    Cell: 302-690-3835

    Fax: 302-831-6243

     

     

     



  • 2.  RE: Time to end TI-84 calculator?

    Posted 07-14-2017 09:15
    Up until several (< 5) years ago I taught college-level statistics as well as a first course in business math and we used Minitab for in-class and home/project work. The school had both Minitab and Excel available and the consensus among the faculty was that Minitab was preferable. The school also tried to get students to purchase an H-P scientific calculator, but few did and noone used it.

    ------------------------------
    Nestor Rohowsky
    President and Principal Consultant
    Integrated Data Consultation Services, Inc.
    ------------------------------



  • 3.  RE: Time to end TI-84 calculator?

    Posted 07-17-2017 09:49

    At UNC in Chapel Hill, we've been using Excel in our largest elementary Stats course for a straightforward reason: our business school requires it of us. A good grade in our course is required for undergraduate admission to the school, and they want to use Excel for many things (e.g. accounting courses etc.), and it is quite important to them that students come out of our course knowing Excel (so they don't have to teach it).
        Having been teaching Excel in such a course for many years, I have a clear understanding of its limitations and problems, and thus a lot of sympathy for the complaints about it.  If I had a choice I would use something more statistics oriented (which is really a different goal from the spreadsheet that is what Excel actually is), perhaps among the options that have been mentioned in this thread.  However, I think that Excel may be getting unfair treatment here.  Jeff Simonoff correctly pointed out years ago some very serious numerical problems, however those have now been fixed.  In my view, an even more serious problem was the organization of functions, but that too has now been quite well fixed.  Both of these major changes were introduced very quietly, but they are quite far reaching, and do overcome many of the early objections in the mentioned "literature".
        The expression of support of hand calculators is quite interesting, even surprising to me.  It shows once again that statistics as a field is extremely broad, with lots of different needs out there.  The reason that I would not have imagined anybody advocating for hand calculators is how I see data sets arriving  for statistical analysis.  In the world in which I currently live data arrive in electronic form (usually an Excel spreadsheet, although I have encountered data sets in Word, almost always with data entry mistakes!).  My understanding is that hand calculators require manual data entry (both tedious, and also prone to errors).  Or is there something I don't know about modern hand calculators going on here?
    Best,
    Steve






  • 4.  RE: Time to end TI-84 calculator?

    Posted 07-14-2017 09:15

     

     

    I would argue strongly against using Excel for statistical analysis.  Much has been written about Excel's limitations and shortcomings in this area.

     

    I recommend something like Minitab, a well-respected product, which is primarily a statistical (and quality control) package.

     

    L. Marlin Eby, Ph.D.

    Professor of Mathematics & Statistics

    Statistics Coordinator

    Editor, Analytic Threads

    Department of Mathematics, Physics, & Statistics

    Messiah College

    One College Avenue

    Suite 3041

    Mechanicsburg, PA 17055-6807

    (717) 796-1800 x6850

     

    http://rs974.pbsrc.com/albums/ae226/caseyfronczek/casey%20fronczek/blue_marlin_venezuela.jpg~c200

     

    Sent from my analog computer (The Antikythera Mechanism)

     

     






  • 5.  RE: Time to end TI-84 calculator?

    Posted 07-17-2017 09:44
    I agree Minitab is a better software than Excel for statistical analysis.  And even more so for advanced analysis.

    But if you want students to understand how things are calculated and why, you have to force them to make a confidence interval by hand.  And Excel can do that, and give you the correct value of t based on a probability level and degrees of freedom.  You can show every step with Excel.  I even suggest to students to make their own templates for problems.

    I am not suggesting use Excel in place of software.  Just Excel in place of the TI-84 calculator.

    ------------------------------
    Thomas Ilvento
    University of Delaware
    ------------------------------



  • 6.  RE: Time to end TI-84 calculator?

    Posted 07-17-2017 09:48

    Many years ago I was involved in a similar discussion; then one important reason for use of calculators in (Norwegian) high schools was that they could be used in exams. If this is no longer a strong reason, I fully agree that time is overdue to replace calculators with better solutions at the high school level. The question is what solutions. I agree with L. Martin Eby and others that spreadsheets should not be recommended for statistical analysis (neither for graphics). If high school students become accustomed to using spreadsheets for statistical analysis, universities will have a hard task to convince them that this is not a good idea at university level. Also, spreadsheets will become heavily used for similar purposes by those who seek employment directly after high school. Also, for statistical software for use at the high school level, the recommendations should not only include proprietary software. Spreadsheets with similar functionality as Microsoft Excel are also available in e.g. Libre Office and Open Office. If more appropriate software is recommended there are many options. I may agree with Beverly L. Woods (another branch in this thread) that programming skills should not be a requirement at the high school level, but I think e. g. a carefully designed R package together with an Rcmdr plug-in package, developed and maintained by highly qualified high school teachers in cooperation with statisticians and programmers, could be a solution. This could be used by high school students without programming skills, and the cost should be a small fraction of license costs for proprietary software.


    ------------------------------
    Tore Wentzel-Larsen
    Norwegian Centre for Violence and Traumatic Stress Studies
    Regional Center for Child and Adolescent Mental Health, Eastern and Southern Norway





  • 7.  RE: Time to end TI-84 calculator?

    Posted 07-14-2017 11:27

    I wouldn't trust Excel in my professional statistical work but as a pedagogical tool it's better than a TI-84. (Primarily because of it's visualization capabilities and ease of access.)

    I would use a TI-84 calculator, if I had one, in my professional work but it wouldn't be my first choice for teaching.


    I haven't used Minitab since the last century so it may combine the best of both worlds; statistical accuracy and visualization capability. JMP may offer similar benefits as well.

    ------------------------------
    David Wilson
    ------------------------------



  • 8.  RE: Time to end TI-84 calculator?

    Posted 07-14-2017 11:27
    I assisted a STEM teacher as an AAAS "senior scientist" volunteer in his AP Statistics class at Washington DC's Cardozo High School last year, and certainly agree with you about the desirability of being able to use Excel. But does Excel have some kind of reasonable math/stat graphing capability now? (I always have found my Excel 2007 version's charting feature favors only business-type presentations and very hard to use for basic graphs of functions, and my one attempt to use my Excel binomial distribution spreadsheet with graph, created on my PC, did not seem to graph easily on the school's Apple computers' version/emulation of Excel !). Also, aren't the TI-84 calculators permitted on the AP test itself? I don't think students could expect to use a laptop while taking such an examination!

    Bruce Murrie
    Retired ED statistician/analyst
    Elementary Statistics junior college instructor in 1967,1973 ;-)

    Sent from my iPhone




  • 9.  RE: Time to end TI-84 calculator?

    Posted 07-14-2017 11:27
    I imagine that you are preaching to the choir here.  The lingering appeal of the calculator comes down to the cost.  There's a non-trivial part of the high school population who can't afford a computer much less software beyond Excel.  School districts and some colleges can't afford labs or site licenses. The latest GAISE College Report (which applies to AP classes as well) once again addresses the disservice we do to students who see nothing but the calculator.  The type of technology used to crunch numbers hardly matters if it is in service of understanding the big ideas of statistics. 

    I would argue that there's a certain flexibility of thinking that we cultivate if students see output from multiple software packages.  The high school classroom is not the place to be learning multiple programs or programming R but a rigid expectation that the p-value should be at the far right of the table or at the bottom of the screen (or even the only thing you are looking for) does not do anything for their workplace/life skills.  My non-traditional students use StatCrunch in the first semester and Excel in the second.  I'm working to include output from JMP, SPSS, and Minitab (because I have access, not because they are popular/best/easiest/etc) into assignments and exams for exactly this purpose.  Maybe someday we'll have a third course that teaches the use of software and programming and recruit some budding statisticians.

    ------------------------------
    Beverly Wood
    Assistant Professor
    Embry-Riddle Aeronautical University
    ------------------------------



  • 10.  RE: Time to end TI-84 calculator?

    Posted 07-17-2017 09:42
    I appreciate the debate about what the ideal solution is because it helps me to be able to better prepare my students for their next steps in education/employment.

    Dealing with the reality is another matter all together. Programs like Excel or R or TI-84 calculators are great but the reality is schools and classrooms are not equitably resourced at the high school level. Some neighboring schools have bring your own devices or provide laptops while my school is wholly underfunded when it comes to resources like these. At least some sources say my district is the most underfunded in America. Check out these 2 articles: Reading, Pa., retains title as America's most poorly funded school district - Education Votes  
    'These Kids Are Just Pawns': The Rising Toll of Inequitable School Funding - NEA Today 
    If tech is required to be prepared for the course then budgets need to be written to allow for that.

    Here's a look inside our school statistics program from my perspective:
    The AP statistics teacher has a small class - around 20 students - and they all are signed out a school provided TI-84 calculator for the year to allow them to prepare for the exam and complete homework assignments. Other AP courses like Calculus have TI-84 calculators as well. Any left over TI-84 calculators go to the honors level classes. Last year was my first year teaching honors statistics - the AP Stats teacher suggested that I teach the class with a TI-30 since I wouldn't have enough TI-84's.  I had about 50 honors statistics students (34 in one class) and was able to get 25 old TI-84 calculators to teach Honors statistics. I had to figure out how to move through the material with partial Ti-84 calculator access in the classroom and no TI-84 at home, or resort to using hand calculations with only a TI-30 (which everyone would have in the classroom and most would have an equivalent version at home). During the year this meant we spent an exorbitant amount of time learning how to do hand calculations so that students could at least do the problems. As seniors, most had never used a TI-84 prior to this year and as a result some students preferred doing the calculations by hand because they said they felt like they had more understanding of what was happening with the numbers. It was extremely frustrating trying to teach a course so critical to their future and that I know opens so many career opportunities with resources that put them in the dark ages compared to students who would be competing with them for jobs down the road.

    Cell phones are the most wide spread resource my students have - though not even all have this. As the course progressed about 5 students were able to download free Wabbitemu versions of the TI-84 on their android devices, which did help. They were doing their 1-2 calculation heavy homework problems in under 5 minutes while the remainder of the class still needed about 40 minutes. My Apple cell phone users were annoyed that they couldn't find an equivalent app. We had developed a level of honesty by the end of the year that during testing android students were permitted to use their cell phone to take the test as long as they kept the device flat on their desk so that all students had access to a TI-84 or equivalent during tests.  If Apple hasn't already, I hope they do come up with a free version of the TI-84 that functions the same as the actual calculator. My students that intend to pursue statistical careers certainly learned the value of the TI-84 and stated they plan to make it their priority purchase for college... so if they don't need it in college PLEASE let them know asap.

    I was fortunate to get regular access to a lap top cart with 30 laptops however they took anywhere from 5-30 minutes to just log in during a 47 minute class period and the version of Excel available on the computer did not permit students to do tasks like making histograms. Only tech administrators can download any needed software. We did use the computers some but not as much as I had hoped. Too many of the students don't have access to computers at all at home. Some share computers with the entire family or have extremely old virus ridden computers. A few used the course as a way to convince their parents to invest in a computer. I learned as the year went on that a handful of my students relied 100% on their cell phone to complete any computer tasks - including writing their research papers for English class. Can you imagine typing even a 2 page paper on a cell phone screen! Still others didn't even have a cell phone so even that wasn't an option. 

    If you are blessed to work with any of my students in the future, know that they had to work very hard and overcome many obstacles to reach you.


    ------------------------------
    Flora Quevedo
    Math Teacher
    Reading High School
    ------------------------------



  • 11.  RE: Time to end TI-84 calculator?

    Posted 07-18-2017 08:09
    I agree that access to Excel is extremely helpful to statistical learning and that the TI-84 almost has a lock on statistical and other mathematical calculations in high schools.  As a high school teacher in the Bronx, I agree with Flora that not only can my students not afford their own laptops software and TI-84s, access is sporadic in the school and students cannot take school owned devices home.  Educational inequity is an important issue here and working to make statistical software and calculators available to all students is a piece of the solution, whether through cell-phone access or through pushing for inexpensive calculators and laptops.  





  • 12.  RE: Time to end TI-84 calculator?

    Posted 07-19-2017 09:15
    I am going to teach Engineer Statistics this Fall.  A large part of the curriculum that I am planning will include an introduction to R.  R has a large learning curve, and I hope to help them over come a large part of that curve before they begin their full time position as entry level engineers.

    Part of my curriculum include explaining why the should stay away from Excel for statistical analysis.

    The commercials software packages are great but can sometimes be an up hill battle with your employer to buy them due to the cost and any possible IT issues.

    ------------------------------
    Shawn Capser
    Student, University of Toledo
    Consultant in Reliability Engineering
    ------------------------------



  • 13.  RE: Time to end TI-84 calculator?

    Posted 07-19-2017 22:50
    I would strongly suggest that you use a software package that is common in industry. Minitab, JMP and Design Expert are some of the mainstays in industry. If you try to include R in your curriculum, you will spend a lot of valuable time teaching a software package that they will have a hard time using in industry. Spend your time dealing with the ideas in the class not how to get the answers with R. 

    As someone that spent about a decade working in industry, trying to get a program that the company has not already decided will be in a standard image will be an issue. Most companies don't let employees downlaod software, no matter how valuable, without a lot of paperwork. 

    Think about it this way.  An IT person will need ot come into your office and borrow your computer to download and install a new piece of software. That takes time and money. To get that software installed, you need ot justify, to the liking of the IT department, why you need that software. So, you need to justify getting that software. Can you justify adding R to your computer given you already have JMP/Minitab and MatLab? What makes R so special that MatLab or JMP/Minitab can't do it? Claiming, "I learned how to use R, not Matlab" is not sufficient. "R does ....." will lead to the next question from IT, "So does..... What makes R special?"

    You also have to deal with sending files to co-workers. If you have R, and they don't, you need to change your format to meet their software.

    R is a fine product. For most of the stats most people do, it is overkill.

    Spend your time wisely teaching the ideas of stats, not the lexicon of R.

    ------------------------------
    Andrew Ekstrom

    Statistician, Chemist, HPC Abuser;-)
    ------------------------------



  • 14.  RE: Time to end TI-84 calculator?

    Posted 07-20-2017 06:59
    I would only like to stress the importance of graphing raw data.
    There cannot be too much emphasis on seeing the shapes of distributions
    before you do any interpretation of statistical tests.
    Especially in intro courses, students may have no experience with the
    fields from which the data come and their intuition may be way off,
    leading to unwise conclusions.
    So, any technology that makes graphing convenient is extremely valuable.




  • 15.  RE: Time to end TI-84 calculator?

    Posted 07-27-2017 04:26

    I can follow the previous posts by Andrew Ekstrom to some extent. In particular he may well be right that the Microsoft R Open implementation is better than standard R for large and numerically demanding problems. Still, I cannot quite agree that it is overkill to teach students who may go to industry jobs to use R. See e. g. http://r4stats.com/articles/popularity/ , I assume that at least some of the jobs included in this overview are in industry. I do not think R is just another statistical program. For instance, does Springer have own book series "use JMP!", "use Minitab!" or "use Design Expert!"? Similarly, Wiley, at http://eu.wiley.com/WileyCDA/Section/id-815382.html lists 25 titles under Statistical software/R, the corresponding figures for other programs listed are 1 for Mathematica, 6 for Matlab, 1 for Maple, 7 for SAS, and 6 for SPSS (similar information at https://www.crcpress.com/ was temporarily unavailable). As to personal experience I have about 9 years experience from a university hospital. R was available for researchers upon request when I started in 2001, and after some years was established as generally available and supported software for all hospitals in Western Norway (together with Stata and SPSS). This is in an environment with strict security measures to protect patient privacy and data integrity. Yes, there was some paperwork involved, but it was worth it. When I started in my present employment in 2010 I got access to R from the outset, the paperwork was not unsurmountable. R is generally available also in some other environments with strict security, an example is http://www.uio.no/english/services/it/research/sensitive-data/index.html . Here, local repositories within the safe environment are created and regularly updated including all packages available under CRAN and Bioconductor. Also I have seen a similar statement for at least one institution using the SAIL Databank (https://saildatabank.com/). Unfortunately, I am not at present able to retrieve the information on this institution. I agree that it is wise to put a strong emphasis on ideas in teaching, and particular software implementations may in some cases take too much time at an early stage. Still, I think implementations in good software programs also has a place in teaching, and experience with working in R may give training in structured work that is valuable also if R is not immediately available in subsequent employments.



    ------------------------------
    [Tore][Wentzel-Larsen][Tore Wentzel-Larsen
    Researcher
    Norwegian Centre for Violence and Traumatic Stress Studies,
    Regional Center for Child and Adolescent Mental Health, Eastern and Southern Norway]
    ------------------------------



  • 16.  RE: Time to end TI-84 calculator?

    Posted 07-27-2017 15:33
    I think the reason you see more titles for R is precisely because it's freeware. The other software packages have a ton of support and documentation provided by their respective companies. The amount of documentation found at https://support.sas.com/documentation/index.html, for example, is daunting.


    _________________________________________
    Gabe Farkas
    gfarkas@gmail.com






  • 17.  RE: Time to end TI-84 calculator?

    Posted 07-17-2017 09:44
    Beverley, I fully appreciate your post!

    ------------------------------
    Thomas Ilvento
    University of Delaware
    ------------------------------



  • 18.  RE: Time to end TI-84 calculator?

    Posted 07-14-2017 11:27
    In a way, the TI-84 is just like an Excel Spreadsheet with its six-numbered Lists (plus 14 more).
    Furthermore, it has the Normal, T, Chi-Square, and F Distributions built into it, as well as Stat 101 interval estimations and significance tests (not to mention one-way ANOVA, which could easily be extended to two-way ANOVA).

    Thus, "Time to end TI-84 calculator?" Absolutely not!
    [These days with students able to afford various other devices, most could get a TI-83/84 (the TI-83 doesn't have the Chi-Square GoF Test but practically everything else), and use it for stat and other things for years to come!]







    ------------------------------
    David Bernklau
    (David Bee on Internet)

    ------------------------------



  • 19.  RE: Time to end TI-84 calculator?

    Posted 07-17-2017 09:44
    Someone else noted that in high schools, it might be a cost thing.  Now everyone taking stat has a laptop.
    In college, or at least at my university, everyone has a laptop.  Our school gives a free version of Office for macs and Pcs.  Everyone has access to Excel.  And I think everyone should know Excel.  The working world uses it a lot.

    But the Ti-84?  I don't think anyone uses it beyond high school.  Maybe I am wrong.  However, I run into a lot of students who have one and completely forget how to use it.

    ------------------------------
    Thomas Ilvento
    University of Delaware
    ------------------------------



  • 20.  RE: Time to end TI-84 calculator?

    Posted 07-17-2017 12:02
    While the TI-83 does not have the Chi-Square GoF Test, programming it is uncomplicated. I did that a couple of years ago so the students did not have to buy a TI-84 when they already had a TI-83. The TI-83 does have both chi-square cdf( and pdf( .

    I do not have the capability of transmitting the program-do not own a TI-83 and retired last year.

    Dennis Robbins




  • 21.  RE: Time to end TI-84 calculator?

    Posted 07-17-2017 09:43
    I should have been clear, so I will make a few points up front:
    • I am focusing on Introductory Statistics.
    • I think every college students should have exposure to Excel - it is used everywhere.  It is incredibly practical and many businesses assume knowledge of Excel.  We want our students to feel very comfortable with Excel - companies tell us even graduate students need to be good with Excel.
    • I am not arguing that Excel is a good statistical software.  I know it has limitations as an analysis tool.  If you want to use software, JMP or Minitab are better tools.  We use JMP at the University of Delaware, and Minitab is also used in labs.
    • However, if you teach introductory statistics and want students to calculate a z-score or solve for the variance, or generate expected frequencies, I think Excel is a better tool.  I am telling my students don't but a calculator, let Excel be your calculator.  It can also give the critical value for t using a built in function, of an F value or ChiSquare.  
    The TI84 seems to have a lock on high school teaching and it seems to me that part of the exam is a test on how to use the calculator.  I have many college students who ask how to use their expensive TI-84 and I say I don't have a clue.   A $10 calculator is fine in my class or I suggest just use Excel as your calculator.

    I am really arguing that Excel has much to offer in teaching statistics if you want students to calculate things by hand.  I say this even if you also use software as part of the course.

    ------------------------------
    Thomas Ilvento
    University of Delaware
    ------------------------------



  • 22.  RE: Time to end TI-84 calculator?

    Posted 07-17-2017 09:44
    As part of the NSF funded Passion-Driven Statistics project, we have begun to regularly use SAS Studio (in the cloud) not only at the college level, but recently, also with our high school student groups.  It is free for educators, requires only access to the internet and can be used on the most inexpensive computers, on any platform and without special assistance from internal IT staff.

    More information describing these experiences is available here
    http://passiondrivenstatistics.com/2016/09/21/gear-up/

    We have also developed some instructional videos for students:

    http://sas.passiondrivenstatistics.com/


    Also, Scarsdale High School (Scarsdale New York) has been successfully using R in portions of their AP statistics course for the past few years. Laura Estersohn is the lead on this and could share experiences. lestersohn@scarsdaleschools.org



    ------------------------------
    Lisa Dierker (ldierker@wesleyan.edu)
    Wesleyan University
    ------------------------------



  • 23.  RE: Time to end TI-84 calculator?

    Posted 07-17-2017 13:33
    Having taught both AP Statistics and now Introductory Statistics at the college level, I can say that testing is still one of the major reasons to use a graphing calculator. Students need access to statistical software during instruction time but their computers (and hence software) can't be used either on my high school class exams, the AP exam or on my college exams.

    I recommend the TI-Nspire over the TI-84 because it addresses many of the above concerns mentioned in earlier posts and functions much more like statistical software. You can import data sets directly (no typing in needed) and it can deal with missing data values. Lists need to have names as in statistical software and the output looks more like what I get with JMP or R-Commander: graphs are appropriately labeled and scaled, etc. You can save data, graphs and data analysis on the Nspire like you can with statistical software. Unfortunately, many high school teachers are so comfortable with the TI-84 that the TI-Nspire is not taking over the market like I feel it should.

    I also have had both my high school and college students use statistical software as part of their course. At the high school level, we had a site license for JMP so my students learned to use that and completed assignments throughout the year using JMP. I use R-Commander at UCI and require its use on about half of my homework assignments. But again, the students can't use their computers for exams so enter the graphing calculator. I do use a lot of computer output on my exams and have students pull appropriate information from it but if I want a student to go through the entire process of inference, they either have to do it by hand (silly in my book) or use a graphing calculator.


    ------------------------------
    Lee Kucera
    UC Irvine, Department of Statistics
    AP Statistics Teacher, Department Chair-Retired
    Capistrano Valley High School
    ------------------------------



  • 24.  RE: Time to end TI-84 calculator?

    Posted 07-18-2017 08:53
    I'm going to go in a completely different direction: Technology beyond a basic calculator (eg TI 30) should be banned in intro stats.  

    Technology use focuses students on getting the right numerical answer from a data set with no real understanding of what that answer means.  The emphasis is on computation rather than understanding

    With computers/calculators students can rapidly compute the mean and median of a reasonably large data set.  But how many students coming out of intro stat appreciate the actual difference in the information provided by the mean and the median?  

    Online videos of statistics lessons seem to follow a pattern:  Show an example or two worked by hand, then jump immediately to computation via technology.  A focus on computation, not understanding.

    Students taking intro stat courses need to understand the statistics they are encountering, not how to compute rapidly and accurately.  
    Sadly statistics instructors have bought into to the claims that ease of computation is the same as depth of understanding.  I see no evidence of that.  All I see is elevated grades among students who are more capable than before of computing answers they do not understand.

    Michael


    ------------------------------
    Michael Granaas
    Univ. of South Dakota
    ------------------------------



  • 25.  RE: Time to end TI-84 calculator?

    Posted 07-19-2017 08:43

    I agree with Michael Granaas. I have never taught high school, but my medical school intro stats class uses relatively little calculation. They learn an unpaired t-test with equal sample sizes as an example of how a typical statistical test works. Likewise how to calculate a CI for a single mean.  I'm much more interested in whether they can tell me what a CI for a mean difference or relative risk MEANS. Can they determine not only statistical significance from it, but also what values of the parameter are ruled out (with 95% confidence) or consistent with the CI, as an aid to clinical significance? Similarly for p-values and hypothesis test results, and so on. Even for basic stats like mean and median, the choice of which to use and the interpretation, especially if they differ, takes as much time in my class as calculations,maybe more. Nobody needs a graphing calculator!


    Ed


    Ed J. Gracely, PhD
    Associate Professor
    Family, Community, & Preventive Medicine

    College of Medicine

    Associate Professor

    Epidemiology and Biostatistics

    Dornsife School of Public Health

    Drexel University
    2900 W. Queen Lane,
    Philadelphia PA, 19129

    Tel: 215.991.8466 
    | Fax: 215.843.6028
    Cell: 609.707.6965

    eg26@drexel.edu (egracely@drexelmed.edu forwards)
    drexelmed.edu  |  drexel.edu/publichealth






  • 26.  RE: Time to end TI-84 calculator?

    Posted 07-26-2017 23:27
    I agree with Ed Gracely and Michael Granaas.

    I am one of those old enough to have had to survive undergraduate years without a calculator of any kind. For long years I carried on about students in Intro Stats who use a calculator to obtain the square of a two-digit number or the square root of 81, for example, or even to do 45–29.


    Far better an approximate answer to the right question, which is often vague, than an exact answer to the wrong question, which can always be made precise.
    The future of data analysis. Annals of Mathematical Statistics 33 (1), (1962), page 13.

    https://en.wikiquote.org/wiki/John_Tukey

    Dennis Robbins




  • 27.  RE: Time to end TI-84 calculator?

    Posted 07-19-2017 08:44
    I personally think it is a waste of time and space condemning programmable calculators as by TI, Sharp, HP, and what not.  It is the teacher's responsibility to teach the students how to compute basic statistical aspects of data before using a calculator, programmable or not. After that, the emphasis should be how to do it right on a calculator.  Use validation of the programs written by a user or the calculator's buttons by hand verifying examples (from good books and websites) before using them.  Let me give an anecdote from my old days when I was a mathematical statistician at the National Cancer Institute, NIH some 40+ years ago.  We had access to two main frame computers at NIH- an IBM 360/370 and a DEC 10.  We were given certain amount of computer time for our research which was not much.  We were also given some simple TI/HP calculators, some of them could have been programmed using key strokes (kind of like machine language programming).  Often we would use these calculators' keys to obtain basic statistics such as mean, SD, etc.  A post doc who  was working with me one day brought a data set and simple statistical parameters for quick examination.  Out of curiosity (the data set was small) I hand calculated them.  My mean matched hers but my SD did not.  Thinking I made a mistake, I recalculated and confirmed what I just found.  I then asked her to show me how she got the SD and she showed me by hitting the key SD.  She did not make a mistake but...  I then hand calculated the SD by not dividing by the df (=n-1), but by n and also repeating by dividing by the df.  Low and behold, the calculator was calculating the population SD by dividing by n.  I then found a key on the calculator that said VAR.  I used that key and took the square root.  Guess what- I now got the correct sampling SD.  The rationale behind it is unknown to me.  The lesson for me was know what you want and know how you are getting it.  Since then, we programmed a TI59 programmable calculator using key strokes even to perform simple nonlinear regression by Newton-Gauss algorithm, factorial ANOVA, multiple linear regression and published some of them for people who did not have access to computers.

    Simply put, these calculators had limitations (less now-a-days) and more often than not, they were not programmed by a statistician (often by engineers dabbling into statistics) but using them correctly, one could do many things handily, including tax returns! Since then, I have seen statisticians writing programs using Lotus 1-2-3 and now its grandchild MS Excel verifying the results using BMDP (Biomedical Data Processing) (remember SAS was in its placental stage then).  I have also seen people complaining about mistakes in their pre-programmed procedures. In any case, instead of throwing these calculates in the ocean (thereby killing fishes and dolphins) put it on your shelves and try to teach your students how to calculate and program statistical procedures. For advanced students (college and university level) I would teach a simple fact of life "SAS is not statistics, it is just a statistical tool". Could be said about any package, not just SAS which is still widely used as a teaching tool The reason is obvious.

    Ajit K. Thakur, Ph.D.
    Retired Statistician





  • 28.  RE: Time to end TI-84 calculator?

    Posted 07-19-2017 08:45
    Michael,
    I agree that computing by hand, what I call getting "down and dirty with the data" is important in an intro stat course.  I tell my students it is useful to have to actually calculate something, but later we would use software.

    My vote is a combination of both,  As to the suggestions of using SAS or R for an intro course, my reaction is why? and good luck.  To really learn the software, even for basic things, is a huge distraction in an intro course.  Of course, you can give them the code, but then what did they learn?

    Finally, my reaction to those that think it is too much to allow students to have resources when taking an exam, except a TI-*4 calculator, my reaction is that the exam becomes in part a test of the calculator.  I just talked with an incoming freshman who messed up her AP Stat exam because she forgot a crucial part of the TI-84.  I believe it was a chi-square test.  That makes no sense to me.  I would much rather allow students cheat sheet of formulas, but they know how to put things together and know the notion of observed minus expected, rather than know how to access the menu in their TI-84.

    ------------------------------
    Thomas Ilvento
    University of Delaware
    ------------------------------



  • 29.  RE: Time to end TI-84 calculator?

    Posted 07-19-2017 23:13
    I agree that "Students taking intro stat courses need to understand the statistics they are encountering, not how to compute rapidly and accurately." But using computers just to calculate is a misuse of computers. The computer needs to be used wisely and carefully, to aid understanding, not just to compute "answers". Use the computer to compute, but then ask questions that require understanding, thought, and careful interpretation. Use computers to do simulations that help students understand concepts. Use computers to produce graphs that aid in understanding concepts. If the textbook you use doesn't have such exercises for students, find another one that does! And, of course, be sure your homework and exams are designed to develop and test for understanding, not  computation.

    ------------------------------
    Martha Smith
    University of Texas
    ------------------------------



  • 30.  RE: Time to end TI-84 calculator?

    Posted 07-19-2017 13:41
    Because smart phones seem to be ubiquitous, perhaps an affordable solution is these two, free, cellphone apps: 1) "Probability Distributions" by Mathew Bognar, Ph.D. University of Iowa, a great, at-your-fingertips resource, and 2) "RealCalc," a nifty scientific calculator that can be configured for RPN if that's your choice (as it is mine).

    ------------------------------
    Charles Annis, P.E.
    Statistical Engineering
    ------------------------------



  • 31.  RE: Time to end TI-84 calculator?

    Posted 07-20-2017 09:54

    You NEVER want smartphones in any type of test environment.  It is too easy to cheat via messaging.  Calculators are the easiest and cheapest way to provide the needed technology for a test environment, especially the AP test. 

     

    By-the-way...cheating is possible on the TI calculators because students can enter notes into the memory.

     

    Barbara Wojciechowski, MS, MA

    Office of Science Policy, Strategic Planning, Analysis, Reporting, and Data (OSPARD)

     

    National Institute on Minority Health and Health Disparities

    National Institutes of Health

    6707 Democracy Blvd

    Bethesda, MD  20892

     

    Email: barbara.wojciechowski@nih.gov

    Phone: 301-451-2417 | www.nimhd.nih.gov

    Connect with NIMHD: Twitter | Facebook | YouTube | Email Updates

    image001.jpg@01D1B662.B6DAD520

     






  • 32.  RE: Time to end TI-84 calculator?

    Posted 07-26-2017 07:55
    HiPer Scientific Calculator does everything RealCalc does and more. I recommend you try it out. 

    Gabe Farkas