Discrete Vs Continous data
Six Sigma – iSixSigma › Forums › Old Forums › General › Discrete Vs Continous data
 This topic has 33 replies, 14 voices, and was last updated 14 years, 11 months ago by fake accrington alert.

AuthorPosts

October 25, 2007 at 5:48 pm #48508
C SureshParticipant@CSuresh Include @CSuresh in your post and this person will
be notified via email.Is Fibonacci series discrete or continuous? Why ? Can anyone clarify.
0October 25, 2007 at 7:41 pm #163879
Dr. Darth aka DarthParticipant@Dr.DarthakaDarth Include @Dr.DarthakaDarth in your post and this person will
be notified via email.For those who don’t know what the Series is here is a quick description of this very simple concept developed around 1200. The first two numbers in the series are one and one. To obtain each number of the series, you simply add the two numbers that came before it. In other words, each number of the series is the sum of the two numbers preceding it. Great for mathematicians tired of masturbation who still need something to occupy themselves but not much use otherwise. If we define continuous data as being infinitely divisable, the Series fails since all numbers in the series are whole and thus discrete. While the series gets very large, it might be possible to justify a consideration of continuous behaving data although the underlying numbers are clearly discrete. Hope that explains it.
0October 25, 2007 at 7:49 pm #163883More than I ever wanted to know about some guy adding numbers. They did not teach the anecdote about mathematicans where I went to school.
0October 25, 2007 at 8:44 pm #163886Discrete
0October 25, 2007 at 11:10 pm #163896What kind of discrete data is it: nominal or ordinal and why? The series of data appears to be continuous to me?
0October 26, 2007 at 1:06 am #163901
Dr. Darth aka DarthParticipant@Dr.DarthakaDarth Include @Dr.DarthakaDarth in your post and this person will
be notified via email.Appears to be is the key. A pound is a continuous measurement since you can subdivide into half a pound or a quarter of a pound or an oz. and so forth. Same for an hour, it can be divided into smaller and smaller units. The Series are all whole numbers and no subdivision takes place, just more and more whole numbers. Thus it remains discrete. End of story. Obviously you are not knowledgeable as to the difference between discrete and continuous nor the different types of discrete. Otherwise you wouldn’t ask if it were nominal. Do a little reading first and see if you can’t figure this out. The answer will come to you quickly.
0October 26, 2007 at 5:09 am #163924How about you taking your own advice and do some reading.
Attribute data is the same at Discrete and there are 3 types of Discrete 1. Binomial 2. Ordinal 3. Nominal
Agreed; weight (such as a pound) is a continuous or variable data type as is time (such as an hour).
Whole numbers have nothing to do with discrete or continuous. I can measure a series in smaller decimal units. Try supporting your response with mathematical support that series data is Discrete.
Darth your respnse was very arrogant, try being more respectful when you respond to others. I just asked a very simple question and you did not answer it.
I do not deserve to be spoken to as if i am some idiot. What comes around goes around0October 26, 2007 at 5:17 am #163926
What language game?Member@Whatlanguagegame? Include @Whatlanguagegame? in your post and this person will
be notified via email.Dr. Darth, help me think that through a little more, because the question is nevertheless puzzling. If I understand the series correctly, the series of numbers is nonreferring. In other words, the series of numbers is selfreferent (you call that masturbation) and not referring to attributes of objects in the empirical world. The assignment of numbers to attributes of objectives is what constitutes measurement. It is in this empirical context of measurement attributes of objects that the numbers are given their meaning as discrete or continuous based on the mathematical operation that are adequate for the type of measurement used to described the object (does exist/does not exist = 0/1, is 10,11, 12 yeards long, etc. When the numbers are not used in the context of measurement as is the case in the series the terms discrete or continuous in the sense that we use them for measurement and statistical analysis are simply not applicable. It’s a different language game altogether.
0October 26, 2007 at 6:30 am #163928
Dr. ScottParticipant@Dr.Scott Include @Dr.Scott in your post and this person will
be notified via email.C Suresh,
It is exponentially ordinal follow a taw or PHI distribution. It is ordinal because each succeeding value is larger than the prior. It is exponential because the succeeding values grow more in value than the previous (that is in a non linear fashion) because each succeeding value is dependent upon the value preceding it. The function is that equal (or similar) to a PHI type function.
Sort of like bees or rats breeding assuming no incest is taking place.
I would like to know what this data represents and how you plan to use it.
Thanks,
Dr. Scott0October 26, 2007 at 6:37 am #163929
Dr. ScottParticipant@Dr.Scott Include @Dr.Scott in your post and this person will
be notified via email.Just to correct some typos:
C Suresh,
It is exponentially ordinal following a tau or PHI distribution. It is ordinal because each succeeding value is larger than the prior. It is exponential because the succeeding values grow more in value than the previous (that is in a non linear fashion) because each succeeding value is dependent upon the value preceding it. The function is that equal (or similar) to a PHI type function.
Sort of like bees or rats breeding assuming no incest or death is taking place in the animals.
I would like to know what this data represents and how you plan to use it.
Thanks,
Dr. Scott0October 26, 2007 at 10:13 am #163949
C SureshParticipant@CSuresh Include @CSuresh in your post and this person will
be notified via email.Thank you Dr. Scott and others for your valuable responses. It was just a query that came up when understanding Discrete and Continous Data.
0October 26, 2007 at 1:11 pm #163958
Dr. Darth aka DarthParticipant@Dr.DarthakaDarth Include @Dr.DarthakaDarth in your post and this person will
be notified via email.Sorry Abel, I stopped reading your response after your first statement since it was incorrect. Back to the drawing board for you. Please read this definition posted on this site and reflect.
http://finance.isixsigma.com/dictionary/Attribute_Data95.htm0October 26, 2007 at 2:49 pm #163963
Ordinal scales and transformatParticipant@Ordinalscalesandtransformat Include @Ordinalscalesandtransformat in your post and this person will
be notified via email.FYI, ordinal scales and permissible transformations:
http://books.google.com/books?id=uek5C3YDU0UC&pg=PA2&lpg=PA2&dq=exponential+transformation+of+ordinal+scales&source=web&ots=JQ_pCOflWr&sig=zWFIdaWmrtZ75DdudDZup4Yn2y0#PPA1,M10October 26, 2007 at 3:16 pm #163967
Queston for Dr. ScottParticipant@QuestonforDr.Scott Include @QuestonforDr.Scott in your post and this person will
be notified via email.A question for Dr. Scott, In your opinion, are we truly dealing with a “measurement scale” here? It seems that the series itself is a series of numbers according to a rule, not the assignment of numbers to attributes of an objecti according to a rule. The series follows an exponential distribution, but they are not anchored in anything observable. Is this series “scalar” in terms of numers theory rather than measurement theory? Which, in my opinion, makes this question so odd and theoretical.
0October 26, 2007 at 3:22 pm #163968
TaylorParticipant@ChadVader Include @ChadVader in your post and this person will
be notified via email.I learned way more than I wanted to on this subject today
Cheers0October 26, 2007 at 4:21 pm #163970Dr. Scott, maybe you could take a look at the comments by Darth and arm him with some increased knowledge in this area, he seems to think that Attribute data and Discrete data are not the same and that Attribute is just binomial. Does not believe discrete cann be classified as ordinal.
0October 26, 2007 at 4:23 pm #163971Thank you Ordinal scales and transformat and excellent link, perhaps Darth can read and understand my question was legit.
0October 26, 2007 at 4:26 pm #163972The definitions on this site that you provided are not valid – I suggest you refer to a valid source as that provided by “Ordinal scales and transformat” or an everyday statistical book if you own one.
0October 26, 2007 at 5:06 pm #163978
Ken FeldmanParticipant@Darth Include @Darth in your post and this person will
be notified via email.I will concede that attribute/discrete and variable/continuous have been used interchangeably. I will continue to argue that a distinction can be made between attribute and discrete although common usage does refer to them as the same. As to the original question, the Series is not continuous data.
0October 26, 2007 at 5:09 pm #163979
Ken FeldmanParticipant@Darth Include @Darth in your post and this person will
be notified via email.Wow, a link to a book cover, now that’s definitive proof but seemed to satisfy Abel so guess that ends this discussion.
0October 26, 2007 at 5:29 pm #163981Agreed, one one ever said it was continuous, just taht it appears as such. The data is Ordinal.
0October 26, 2007 at 5:30 pm #163983
Chris SeiderParticipant@cseider Include @cseider in your post and this person will
be notified via email.Darth,
Where were you during the last discussion on discrete and attribute? There was too much arguing during that session also. I stated that discrete is typically numerical and attribute is word or letter based but both are categories.
Of course, I wonder why some worry about these distinctions. Let’s solve the business or personal problems and not get hung up on semantics.
Good to see you posting more again.0October 26, 2007 at 5:34 pm #163985I’ll be honest with you, i skimmed it, but saw enough references to conclude it worthy
0October 26, 2007 at 5:37 pm #163987
Ordinal scale and transformatParticipant@Ordinalscaleandtransformat Include @Ordinalscaleandtransformat in your post and this person will
be notified via email.Dr. Darth, this is a somewhat funny site. At times, clicking on the link will allow you to access the relevant pages (2 7) at other times it will simply lead you to the cover page. The first seven pages of Cliff’s book provide an excellent overview of the issues and controversies surrounding the concept of “scaling” and “scaling levels” (with very well documented further references). The five pages cover both Steven’s theory (1951) as well as the subsequent development of his concepts by Luce and Tukey (1964) via conjoint measurement. It also covers Dr. Scott’s approach of determining the scale by fitting models which goes back to Gulliksen (1946). As always with these very conceptual issues a simple right or wrong answer is difficult to establish.
0October 26, 2007 at 7:50 pm #163996
Ken FeldmanParticipant@Darth Include @Darth in your post and this person will
be notified via email.Great to hear from you again Seider, a blast from the past. That is the point I was trying to make but it is splitting hairs. I agree that solving personal issues on this site is much more productive than dealing with the technical crap :).
0October 26, 2007 at 8:16 pm #164001
Ordinal scale and transformatParticipant@Ordinalscaleandtransformat Include @Ordinalscaleandtransformat in your post and this person will
be notified via email.You really have to give the Six Sigma guys credit for always winning their game. They want to look academic, they put their doctor hat on and throw out big words. They get entangled in the complexity of their words, they pretend to pull up their sleeves. Suresh gets blasted for being too theoretical, Abel for being too unknowledgeable. In all cases, Darth and Seider win the game for appearing to be scientific pragmatists and pragmatic scientists all in one whithout ever solving the problem. That’s the witchcraft of modern day consultants. And it works and pays the bills!
0October 26, 2007 at 8:18 pm #164002
Chris SeiderParticipant@cseider Include @cseider in your post and this person will
be notified via email.Darth,
Wow, you may have just opened up a new string of posts! Some people seem to use this site as a forum for cleansing….
It was nice earlier this month to see the kind notes about the individual who passed away….kudos to those who spent the time to let others know the bad news of someone passing away. I did not the individual, Reigle Stewart.0October 26, 2007 at 8:22 pm #164003
Chris SeiderParticipant@cseider Include @cseider in your post and this person will
be notified via email.err… “I did not know the individual”.
0October 26, 2007 at 8:46 pm #164005
Ordinal scales and transformatParticipant@Ordinalscalesandtransformat Include @Ordinalscalesandtransformat in your post and this person will
be notified via email.not to fear, Darth can handle it.
0October 26, 2007 at 10:00 pm #164010
Dr. ScottParticipant@Dr.Scott Include @Dr.Scott in your post and this person will
be notified via email.There is a lot of merit to what you are saying. My only thought might be that as errors occur earlier in the process, then they exponentially lead to more errors. But I agree, it seems to be more a “prediction” of what the end product might be, rather than any specific measure of a quality.
I still am not sure what this measure is being used for, but then I havent read the entire thread yet.
Regards,
Dr. Scott0October 26, 2007 at 11:05 pm #164011
Dr. ScottParticipant@Dr.Scott Include @Dr.Scott in your post and this person will
be notified via email.Abel,
Bottom line is there are two large categories of data: discrete and continuous.
Both can be ordinal in nature (in fact before data can be considered continuous it must be ordinal). However, discrete data can be ordinal as well, such as counts. 4 is larger than 3 which is larger than two if you are counting defects. Though ordinal, it is still discrete.
The data that C Suresh is referencing is discrete and ordinal. It is discrete because it is ultimately based on counts. It is ordinal because one number can be said to be greater or lesser than another. Furthermore, it is exponential due to the function behind it (Fibonacci function). But it not continuous cause any value between each resulting number has no real meaning (like 1,000.435 bees). What would the .435th bee look like?
Another type of discrete data is attribute in nature or binomial; does the “thing” have the attribute or not (e.g., 0 for no or 1 for yes), but it is not ordinal (only descriptive). That is, does the object have the attribute or not? Discrete data can also be categorical, that is have more than two options such as A, B, C, or D. An example of this world be race or hair color. Such a measure is discrete, categorical, but not ordinal.
And finally, discrete data can be ordinal and in some cases assumed to be interval in nature. An example of this would be a 17 satisfaction scale. If the interval assumption can be accepted, then averages and comparisons can have much meaning and use. That is, though technically discrete, in certain cases it can be treated as continuous.
Hope this helps all,
Dr. Scott
0October 26, 2007 at 11:06 pm #164012
fake accrington alertParticipant@fakeaccringtonalert Include @fakeaccringtonalert in your post and this person will
be notified via email.Where is the smart “Andy Urguhart”?
0October 27, 2007 at 12:05 am #164014
Ordinal scale and transformatParticipant@Ordinalscaleandtransformat Include @Ordinalscaleandtransformat in your post and this person will
be notified via email.The categorization is an excellent review of the “received view”. The brutal relaity is that underlying the classification schemes of discrete/continuous and nominal/ordinal/interval/ratio are two historical strands of measurement theories: The classical Eucledian and the modern version of Steven. So far, all attempts at reconciling the two views have not resulted in universal consensus. But this is not a matter of Six Sigma, but of measurement theory.
0October 27, 2007 at 9:05 am #164025
fake accrington alertParticipant@fakeaccringtonalert Include @fakeaccringtonalert in your post and this person will
be notified via email.It is becoming part of the comprehensive SS
0 
AuthorPosts
The forum ‘General’ is closed to new topics and replies.