Sample Chat Room Meeting with Student

Archive for 200410PSY36175021435 on Jan 25, 2004
--------------------------------------------------------------------------------
KS has entered. [ 04:54:11 PM ]
Sally Kuhlenschmidt has entered. [ 04:54:14 PM ]
KS > Hello, everyon. It's KS
Sally Kuhlenschmidt > Hi KS
KS > I signed on once and then got off and now I am on again so I don't know if anyone saw my first message
Sally Kuhlenschmidt > I can see a "Hello everyon
Sally Kuhlenschmidt > But I bet it is in the archives.
KS > ok thanks
Sally Kuhlenschmidt > I think it is just you and me.
KS > alright
Sally Kuhlenschmidt > What we say gets saved in the archives for others to see.
Sally Kuhlenschmidt > Just so you know.
Sally Kuhlenschmidt > Have you ever done chat?
KS > I was wondering if any of the any of the case studies will be on the test
Sally Kuhlenschmidt > What could be?
Sally Kuhlenschmidt > They're speculative
KS > no I have never been in a chat before and I am a very slow typer
Sally Kuhlenschmidt > The principles they pull out will be-- but not directly.
KS > ok
Sally Kuhlenschmidt > I'm a very fast typist (as you have no doubt noticed)
Sally Kuhlenschmidt > Don't worry about typing over the top of me-- just send your message.
KS > will you post a study guide for the test
Sally Kuhlenschmidt > Conversations can get a little out of synch in here-- you'll get used to it
Sally Kuhlenschmidt > No, I don't think I will-- I'm giving you a practice exam and plenty of chances to ask questions. I'm not sure what additional value it would provide.
KS > Is the practice test similar to our fist test
Sally Kuhlenschmidt > It's roughly similar-- I didn't try to make it parallel
KS > ok.
Sally Kuhlenschmidt > I don't think you can infer a score from it to the exam...but you can see the types of questions and my instructions.
KS > Do you do key terms from the chapter
Sally Kuhlenschmidt > I did modify the instructions a tiny bit
Sally Kuhlenschmidt > Figured out how to let you write justifications within the test.
Sally Kuhlenschmidt > Key terms?
Sally Kuhlenschmidt > Isn't that just another name for a study guide?
KS > key terms like efficiency versus equity
Sally Kuhlenschmidt > Sounds like you've already spotted them.
Sally Kuhlenschmidt > Let me give you a secret about study guides....
KS > at the end of every chapter they have key terms like definitions
Sally Kuhlenschmidt > All the instructor does is go through the chapter and write down all the bold words and the headings.
Sally Kuhlenschmidt > and the words at the end of the chapter
Sally Kuhlenschmidt > So why don't you gather those yourself?
KS > ok thanks
Sally Kuhlenschmidt > The virtue of doing it yourself is that you have to think about it....
Sally Kuhlenschmidt > and that thinking is the studying.
Sally Kuhlenschmidt > (By the way a ... at the end of a sentence means more to come.)
KS > I was hoping other poeple would join and ask questions that did not occur to me
Sally Kuhlenschmidt > Doesn't look like they will. Might as well go with your questions over the material.
KS > do I need to turn in any papers for formulas or will they be on the computer with the tests
Sally Kuhlenschmidt > I don't understand "turn in any papers for formulas"?
KS > loose leaf papers to write down standard deviation
Sally Kuhlenschmidt > not an option. On the plus side, I don't often require calculations...
KS > good
Sally Kuhlenschmidt > On the tough side, you have to really understand. Mechanical calcuations aren't what I'm looking for.
Sally Kuhlenschmidt > I'm wanting to know, Do you know what a standard deviation tells you?
Sally Kuhlenschmidt > Why would you calculate a mean as opposed to a mode?
Sally Kuhlenschmidt > Do you see the difference?
KS > standard deviation is how close a nuber is to the mean
Sally Kuhlenschmidt > roughly
KS > mean is the average number of scores
Sally Kuhlenschmidt > how is mean different from mode which can also be described as an average?
KS > mode is similar to mean but not as exact
Sally Kuhlenschmidt > Nope
KS > mode is the most frequently occuring value in the set of scores
Sally Kuhlenschmidt > Correct ...but more importantly, they are different ways of summarizing the same data.
KS > I got that off of your web site for chapter 3
Sally Kuhlenschmidt > One isn't "better" than the other except with regard to the starting purpose of calculating them. (Not my website-- I sent you to another guys)
Sally Kuhlenschmidt > Need to revisit your standard deviation understanding too-- not precise enough.
Sally Kuhlenschmidt > in the ball park but not quite there.
KS > I am sorry I did not have many questions ready for the study sesson
Sally Kuhlenschmidt > Here's a good question (shifting gears) when would you use a mode instead of a mean?
Sally Kuhlenschmidt > Contemplate the different characteristics of mean vs mode.
KS > if it is bimodal
KS > two values occur more frequently
Sally Kuhlenschmidt > Then a mode might be a more accurate description of the data-- yes.
KS > ok
Sally Kuhlenschmidt > Or, if one must have a single number...a mean might be better...
KS > yes
Sally Kuhlenschmidt > oops! there isn't a single answer, turns out it depends on the starting assumptions.
KS > ok I am lost
Sally Kuhlenschmidt > That's the way you want to think about stat in this class- what are the starting needs of the situation.
KS > ok
Sally Kuhlenschmidt > Unlike 201/210 Tests and Measurement has one leg always firmly in the practical world.
Sally Kuhlenschmidt > 201/210 has to get very philosophical but 361 type folk are testing for a purpose.
KS > this class doesn't?
Sally Kuhlenschmidt > doesn't what?
KS > have a leg in the practical world
Sally Kuhlenschmidt > Somehow that came out opposite..let's start over and let me rephrase that.
Sally Kuhlenschmidt > 361 always has one leg in the practical world, but 201/210 can go off in flights of pure analysis/mathematics.
KS > ok that sounds better
Sally Kuhlenschmidt > With one leg in the practical world, one must always ask what are the starting assumptions of the situation.
Sally Kuhlenschmidt > Sometimes describing a sample with 2 modes is more useful or meaningful than describing it with a mean and sometimes the opposite is true.
Sally Kuhlenschmidt > Therefore, it is dangerous to think you can find an answer that will fit in every situation.
KS > so one is never really better
Sally Kuhlenschmidt > That's where it gets tricky...
Sally Kuhlenschmidt > In a particular situation, one is best.
Sally Kuhlenschmidt > Let me think of an example...
Sally Kuhlenschmidt > or an analogy...
Sally Kuhlenschmidt > You are planning a meal. what entree should you serve? hot dogs or filet mignon?
Sally Kuhlenschmidt > If you have 12 6 year olds-- hot dogs.
Sally Kuhlenschmidt > If you have a date coming over, filet mignon
Sally Kuhlenschmidt > Make sense?
KS > yes
Sally Kuhlenschmidt > So context is very important.
KS > ok
Sally Kuhlenschmidt > That's why you have to understand the functions of the various statistics, rather than memorize formula.
KS > ok
Sally Kuhlenschmidt > (there's a little bit of memorizing here and there but mostly the harder stuff of understanding).
Sally Kuhlenschmidt > It's a different way of studying than for 201/210.
KS > seems so
Sally Kuhlenschmidt > You'll get it-- just keep at it.
Sally Kuhlenschmidt > It's important to practice application to get it-- that's why I have the case studies to give you several contexts.
Sally Kuhlenschmidt > And to see how the answer "changes" as the context changes.
Sally Kuhlenschmidt > And that's why I can say the cases aren't on the exam in detail but they are in "form" or "principle."
KS > is it really fair to compare scores with other people to make an assumption
Sally Kuhlenschmidt > Ooh complex question and a cool one.
Sally Kuhlenschmidt > Let's dissect it first to make sure we understand the context.
KS > for example 2 people take the same test and I compare their scores
Sally Kuhlenschmidt > First, let's set aside fairness a minute because that's a hard concept.
KS > ok
Sally Kuhlenschmidt > That brings us to issue of comparing scores, specifically to other people.. how is that done?
Sally Kuhlenschmidt > Well, that's a few chapters this term
Sally Kuhlenschmidt > Any measurement is a comparison-- eg. to a yardstick as the standard. That's the whole standard of comparison thing.
KS > does comparing really give performance level
Sally Kuhlenschmidt > Depends on how you define performance...
Sally Kuhlenschmidt > or what question you are trying to answer.
Sally Kuhlenschmidt > If you are asking, who is the best runner, then yes
Sally Kuhlenschmidt > comparing the person to others is implicit in the question-- do you see?
KS > yes
Sally Kuhlenschmidt > If you are asking, how fast can John go, then a stopwatch can answer that.
Sally Kuhlenschmidt > It's a comparison to an external standard
KS > but if the slower runner had a bad day but is usually fast how accurate is that test
Sally Kuhlenschmidt > It's only one sample...
Sally Kuhlenschmidt > remember from chapter 1 that every test is just a sample.
KS > yes
Sally Kuhlenschmidt > You may have to take multiple samples to increase your confidence in the conclusions.
KS > ok
Sally Kuhlenschmidt > Or have to carefully set up the circumstances of testing to maximize the probability of getting a good sample.
Sally Kuhlenschmidt > The whole steroids controversy is trying to make the tests of speed reliable.
KS > like control the variables
Sally Kuhlenschmidt > Yes, exactly
Sally Kuhlenschmidt > In that sense, 361 is like 201/210-- you are doing the research methods, just extending them.
Sally Kuhlenschmidt > Back to comparisons-- there is a 3rd type of comparison...
KS > ok
Sally Kuhlenschmidt > What if you ask, Is John running faster today than he did a year ago? That's relative to self.
KS > yes
Sally Kuhlenschmidt > So there are 3 comparisons: relative to others, relative to an external criteria, and relative to self.
Sally Kuhlenschmidt > I'll ask you to try that out on your webpage.
KS > ok
Sally Kuhlenschmidt > So the answer to your question about is it fair to compare to other people...
Sally Kuhlenschmidt > requires first finding out what the original question was. If it is a relative to external criteria or relative to self question-- then the answer is no.
Sally Kuhlenschmidt > If it is a relative to other people question, then the answer is maybe...
Sally Kuhlenschmidt > which brings us to what is fair?
Sally Kuhlenschmidt > Lots of factors figure into what is fair.
Sally Kuhlenschmidt > including values (which science can't address).
KS > besides assigning numbers to comparisons is there any other way to evaluate behavior
Sally Kuhlenschmidt > Yes-- you can flip a coin
KS > besides assigning numbers is there any other way to evaluate behavior
Sally Kuhlenschmidt > you can have a mystical experience and pronounce a judgement
Sally Kuhlenschmidt > (I'm not being flippant--these are other ways)
KS > judgement is till measurement
Sally Kuhlenschmidt > you can use "judgement" which means going with a feeling
Sally Kuhlenschmidt > Yes, judgement is measurement, and pretty poor
Sally Kuhlenschmidt > No accountability
Sally Kuhlenschmidt > no reliability
Sally Kuhlenschmidt > questionable validity
Sally Kuhlenschmidt > (and that's not my opinion, that's from years of research)
KS > ok
Sally Kuhlenschmidt > On the other hand, there are lots of real world situations where you can only use judgement.
Sally Kuhlenschmidt > Can you think of any other way to evaluate people?
KS > is there ever an absolut zero in statistics
KS > i can evaluate a person by how they a re dressed
Sally Kuhlenschmidt > First judgement, then back to zero
Sally Kuhlenschmidt > That evaluation of how they're dressed-- what does it involve?
KS > i judge their clothes
Sally Kuhlenschmidt > Ahh-- judgement
Sally Kuhlenschmidt > What did I say about judgement?
KS > any interprete it to how I dress and why I would dress like that
KS > it is a poor measurement
Sally Kuhlenschmidt > Yes, poor...lots of variables in why a person dresses in a particular way.
Sally Kuhlenschmidt > So no other way to evaluate performance-- chance, judgement, assigning numbers
Sally Kuhlenschmidt > Now, is there an absolute zero in statistics...
Sally Kuhlenschmidt > First, statistics is math formulas and assumptions, it has the assumptions of regular math which assumes zero (actually, it was India that came up with the idea of zero).
Sally Kuhlenschmidt > Math got along with the concept of zero for millenia...but it makes it easier to have it.)
Sally Kuhlenschmidt > I think maybe you are asking something else
KS > ok
Sally Kuhlenschmidt > Perhaps something about the Scales of Measurement?
KS > yes
Sally Kuhlenschmidt > go one
Sally Kuhlenschmidt > go on
KS > why are there different names ordinal nominal
Sally Kuhlenschmidt > This can be a hard one to get your mind around until you do, then it's like a light switch.
KS > ok
Sally Kuhlenschmidt > For most of your life you assumed a number was a number.
Sally Kuhlenschmidt > Not terribly interesting.
KS > yes
Sally Kuhlenschmidt > Turns out, something may be labelled a 2, but that 2 is different from this 2 in a fundamental way...
Sally Kuhlenschmidt > context again.
KS > ok
Sally Kuhlenschmidt > Here's an easy contrast. The 2 on a football jersey-- does it carry the same information as the 2 on a thermometer?
Sally Kuhlenschmidt > What do you know from the 2 on the football jersey?
KS > no
KS > nothing
KS > mabe a name
Sally Kuhlenschmidt > Ahh-- a name...and what is the function of a name?
KS > not much
Sally Kuhlenschmidt > Why is it funny that George Forman named all his kids George Forman (even the girls)
KS > mabey performance by that person
Sally Kuhlenschmidt > Not performance
KS > i don'st know
Sally Kuhlenschmidt > What happens when Mrs. Forman yells for George?
KS > they all come
Sally Kuhlenschmidt > Exactly-- so the name isn't working like it's supposed to...
KS > right
Sally Kuhlenschmidt > What happens in your house when your mom yells KS?
KS > i run the other way
Sally Kuhlenschmidt > :0) yes, just you, not anyone else
KS > she only uses that name when i do something wrong
Sally Kuhlenschmidt > So the name distinguishes a unique individual...just as the 2 on the jersey does.
KS > yes
Sally Kuhlenschmidt > YOu don't have 2 "2's" on the football team. That's the nominal level of measurement.
KS > but not order
Sally Kuhlenschmidt > nominal means name. Correct--not order.
KS > ok
Sally Kuhlenschmidt > jersey 2 isn't better than jersey10
Sally Kuhlenschmidt > However, if we have a foot race, is 2nd better than 10th?
KS > that is rank
Sally Kuhlenschmidt > yes-- it is a 2 that shows order or ordinal
Sally Kuhlenschmidt > Now we have 2 "2's" but they don't have the same characteristics.
Sally Kuhlenschmidt > One can only uniquely identify a person like a name (nominal)
Sally Kuhlenschmidt > THe other is both a unique identifier and gives order
KS > interval
Sally Kuhlenschmidt > In that race, do you know how much distance was between 2nd and 3rd vs 10th and 11th?
KS > no
KS > unless they give times
Sally Kuhlenschmidt > correct Ordinal scales can't tell the size of the interval between the numbers on them.
Sally Kuhlenschmidt > but if you move to time...
KS > it is interval
Sally Kuhlenschmidt > do you know the interval size between 1 and 2 minutes vs 10 and 11 minutes?
KS > yes
Sally Kuhlenschmidt > Correct-- it's 1 minute, exactly the same.
Sally Kuhlenschmidt > Now "2" has a new bit of information-- it identifies a unique time period and an order ...
KS > ok
Sally Kuhlenschmidt > and it also, because you know the interval, tells how big a difference there was in 2 vs 10
KS > yes
Sally Kuhlenschmidt > Just from knowing a) the scale of measurement adn b) the number
Sally Kuhlenschmidt > Now-- I've misled you.
KS > how
Sally Kuhlenschmidt > Time is actually at the ratio level of measurement.
Sally Kuhlenschmidt > Because there is a true concept of zero with it
KS > because of zero
Sally Kuhlenschmidt > yes correct
KS > ok
Sally Kuhlenschmidt > You can have zero time before the race starts.
KS > yes
Sally Kuhlenschmidt > Bruni and I pick out measures and debate whether they are interval or ratio
Sally Kuhlenschmidt > sad but true.
KS > lool
KS > lol
Sally Kuhlenschmidt > It all depends on the context (again!) of the measure.
KS > ok
Sally Kuhlenschmidt > He and I go around on weight. In space there is zero weight...but can one really have zero weight on earth?
KS > so is ratio not a good measure or is interval a poor measure
Sally Kuhlenschmidt > It depends on...what?
KS > no zero weith on earth
Sally Kuhlenschmidt > context!
Sally Kuhlenschmidt > :0)
KS > the context is what it depends on
Sally Kuhlenschmidt > What is the question you are asking
Sally Kuhlenschmidt > very good
Sally Kuhlenschmidt > If I'm concerned with weight on earth, then I can make a case for it being an interval scale.
Sally Kuhlenschmidt > Once we go off planet, then ratio
KS > ok
Sally Kuhlenschmidt > Although some physicists would argue that as long as you are near any body, there is gravitational pull.
Sally Kuhlenschmidt > PRobably more than you wanted to know.
KS > really
Sally Kuhlenschmidt > Most psychological measures are ordinal, but you do encounter nominal and the occasional interval
Sally Kuhlenschmidt > Physiologists work with ratio
Sally Kuhlenschmidt > This is important because of the inferences you can make from the measures and the types of statistics that are appropriate.
Sally Kuhlenschmidt > depend on the scale of measurement.
Sally Kuhlenschmidt > Here's a practical example.
Sally Kuhlenschmidt > If you ever hear someone say "John is twice as smart as Bill" you can mock them with my permission.
Sally Kuhlenschmidt > KNow why?
Sally Kuhlenschmidt > This is tricky
KS > why
Sally Kuhlenschmidt > tricky until you know about the scales of measurement...
KS > is bill below average
Sally Kuhlenschmidt > "Twice" is a statement of a ratio (2John/1Bill)
Sally Kuhlenschmidt > Intelligence measures are ordinal measures.
Sally Kuhlenschmidt > You can't do meaningful ratios with ordinal measures. It's a nonsense statement.
Sally Kuhlenschmidt > LIke saying the number 2 jersey is better than number 10-- doesn't make sense.
KS > so you do not compare intelligence
Sally Kuhlenschmidt > Comparison isn't a ratio.
Sally Kuhlenschmidt > You can compare-- more than or less than
KS > how is the statement not a comparison
Sally Kuhlenschmidt > Just can't say twice or half
KS > ok
KS > why
Sally Kuhlenschmidt > why which?
KS > twice or half
Sally Kuhlenschmidt > Twice and half are ratios-- they require having a meaningful zero. (must be able to have a zero concept to do ratios
Sally Kuhlenschmidt > because some division produces zero
KS > more or less does not have a zero?
Sally Kuhlenschmidt > (that's the easy way to put it).
Sally Kuhlenschmidt > Correct, you don't need a zero to say there is more of something. For example, you can say 78 Fahrenheit is hotter than 32 Fahrenheit...
KS > i did nto know this was so complicated
Sally Kuhlenschmidt > Fahrenheit is an ordinal scale-- 0 on it doesn't mean anything.
Sally Kuhlenschmidt > real
KS > right
KS > absolut zero is calven
Sally Kuhlenschmidt > One can confuse oneself-- you just have to really watch your logic. ..
Sally Kuhlenschmidt > Right Kelvin has a meaningful zero--when all movement of all molecules stop.
Sally Kuhlenschmidt > To figure out the scale, start with asking if the measure number has nominal property...
Sally Kuhlenschmidt > then check on ordinal...if yes, then I ask if it has a zero-- if not, it's interval. If it does, then it's ratio.
KS > niminal to classify and identify
Sally Kuhlenschmidt > Nominal is the name-- it identifies uniquely or classifies.
Sally Kuhlenschmidt > They all have their purposes-- none better than the other.
KS > ok
Sally Kuhlenschmidt > Classification is the first step of science. Nothing else happens without that.
KS > ok
Sally Kuhlenschmidt > Then researchers work on relationships among the concepts-- often order concepts get worked on.
Sally Kuhlenschmidt > For example, with tornados-- first they try to classify-- then decide if classifying by magnitude is meaningful.
KS > ok
Sally Kuhlenschmidt > Then they try to figure out if there are steps between the levels that relate to meaningful criteria or order.
Sally Kuhlenschmidt > They fold in each of the concepts of the scales in turn--working toward the more complex ratio.
Sally Kuhlenschmidt > How you doing?
KS > ok
KS > is there anything that sums it up this easy in the book
KS > or a chart
KS > or a formula
Sally Kuhlenschmidt > Well, I think this is one of those concepts you just have to approach from multiple directions until you find what works for you.
KS > ok
Sally Kuhlenschmidt > It would be a good idea for you to construct such a table for yourself.
KS > good idea
Sally Kuhlenschmidt > The construction would be good practice and it would be maximally meaningful to you.
Sally Kuhlenschmidt > If you want to send me a copy of your table, I"ll proof it.
KS > ok thanks
Sally Kuhlenschmidt > I have an extra credit task (Under assignments) which involves applying the scales to a hobby).
Sally Kuhlenschmidt > If you have a question about a statistic, you can draw it up above...click on the pencil shape and draw...
KS > ok
KS > i do not have any right now i am over loaded with this
Sally Kuhlenschmidt > Not suprised.
Sally Kuhlenschmidt > Chat can be intense.
Sally Kuhlenschmidt > We covered some pretty deep topics.
Sally Kuhlenschmidt > If you are full we can stop.
KS > do you have any more explanaitions that you think might help
KS > i am open to anything
Sally Kuhlenschmidt > hmm
KS > or we could stop and wait for tuesday
Sally Kuhlenschmidt > Without knowing more about your background, hard to say...
KS > ok
Sally Kuhlenschmidt > if you do some reading between now and Tuesday you can keep a list of other questions you might have...
Sally Kuhlenschmidt > not just in the stat but anything in the chpaters.
KS > ok that is a good idea thanks ffor your time it was a great help
KS > see you teusday
Sally Kuhlenschmidt > Glad it was worthwhile...
Sally Kuhlenschmidt > It gets archived so you can look back at it.
Sally Kuhlenschmidt > Until Tuesday-- I'll wait until you leave.
KS > great
KS has left. [ 06:02:22 PM ]
Sally Kuhlenschmidt has left. [ 06:02:27 PM ]

Contact the author with comments or questions about this site by following the directions at this page (which will open in a new window.)