Homework 2.25 to 2.59
Exercises 2.25 to 2.59 are due on Thursday - you can post a comment if you have any questions. If you cannot get in sign a password - just check the anonymous button to make a comment.
posted by maclin_uni @ 8:03 AM
This is a site for our advanced (graduate) statistics class...
Exercises 2.25 to 2.59 are due on Thursday - you can post a comment if you have any questions. If you cannot get in sign a password - just check the anonymous button to make a comment.
posted by maclin_uni @ 8:03 AM
13 Comments:
Hi Dr. MacLin,
I'm working on the homework assignment and do not know how to go about
questions 2.44 and 2.45. These questions ask for what percentage of
scores fall within plus or minus two standard deviations from the
mean. I know how to figure it out by hand, but was wondering how to go
about it with SPSS.
2.44 asks you to find the percentage of scores falling within plus and minus 2 stdevs. Using the data from 2.1
Here is one way.
1) enter the data and calculate z-scores using the procedure Descriptives. Save the z-scores
2) run the procedure Frequencies on the new z-scores and check the cumulative percentage column on the right. You should be able to subtract the percent falling below and the percent falling above…
3) for the data I entered in table 2.1 (which was rather fast – so don’t use my numbers here unless they match yours) I have a value of -2.01 which falls below the desired range of scores and I have a 2.6 and 4.7 which fall above the desired range. Based on my cumulative percentages, each of these three scores account for 2.1% of the data. As such I would conclude that 6.3% fall outside the range leaving 93.7% to fall within.
Hope that helps,
--Dr. M
Here is another way you could do 2.44 using the recode manipulation found under Transform > Recode
1) after you create your z-scores you can recode your z scores (into a new variable) into 0 for scores that range lowest to -2, 1, for scores that range from -2 to 2, and 0 for scores that range above 2
2) the do a frequency on the new recoded variable
Hi Dr. MacLin,
I'm working on 2.30 for TR homework.
I'm trying to graph the cases by year in Excel. However, excel
refuses to recognize the years as categories. So whenever I try to
graph it it puts a line graph of the case numbers as well as a line of
the years. More simply, how do I put the years as my Y-axis?
thanks
Here is how I did it
1) Using Excel I entered the years into column A and the frequencies into column B
2) I selected the values in col B only to make the line graph
3) From the menu I went Insert>Chart to get the chart wizard
4) I chose the line graph – the wizard went to step two
5) At this point I need to add the valsues for the Years into the x-axis
6) I do this by clicking on the series tab near the top of the chart wizard
7) If you look at the bottom you now see “Category X axis labels” it wants you to define a range of numbers. To do this click on the little icon to the far right just after the text box
8) When you click on it a thin dialog box will appear – ignore this for now and select the range of numbers in column A (just like you selected the numbers in B to start the graph). This will tell Excel exactly which numbers you want for the axis labels and put that information into the thin dialog box.
9) You can see where the numbers have been added to the thin dialog box – just to the right of that is another small icon/button. Click on that and you will be returned to the chart wizard
10) You should see that your values have been added – now continue with the other steps.
Take care,
--Dr. M
I'm having some trouble with the homework problems 2.34 and 2.49. Has anyone had any success with these yet that could help me out? Thanks!
Has anyone had any success deciphering the data for 2.55? It is difficult to find the errors when I don't understand what each column means. Any help??
Exercise 2.34 calls for a unimodal distribution with equal mean and median, but with a different mode. I made this data set with a mode of 1 and a mean and median of 5. 1
1
1
1
1
3
4
5
6
7
8
9
10
11
7
Exercise 2.49 basically is asking you to take the data set you created for 2.46 and convert the data to z-scores....
Comment re 2.55: Has anyone had any success deciphering the data for 2.55? It is difficult to find the errors when I don't understand what each column means. Any help??
Response: I found the data sets at http://www.uvm.edu/~dhowell/methods/DataFiles/DataSets.html
(note: if you are having trouble finding Howell’s data sets try the link above)
There are two data sets (Cancer and Bad Cancer).
2.55 tells you to go to the Appendix Computer Exercises to see how the variables are described (this starts on page 666). This should help you better understand what each column means.
RE 2.58:
I am wondering how to do number 2.58. Are we supposed to be creating 2 boxplots or just one. Also I am having trouble getting the URL to work, I think I found it another way though. What data exactly are we supposed to include in the boxplots?
Response: to me it sounds like you will need two boxplots. You need to use the data from table 2.1 which is organized by the number of letters and the yes \ no response (Howell tends to refer to this as the positive and negative instances. He is asking you to ignore the number of letters and just look at the yes / no responses to see if it takes more time to make a yes response or a no response. If you make two boxplots you should be able to come to a conclusion about this.
Hope that helps,
--Dr. M
Do we really have to make 17 descriptives tables for each of the variables in 2.55? There are 17 variables and it says to make one for each variable.
Does it matter how much work we show? I'm mainly talking about larger equations that we need to do through spss. For example problems 2.41 and 2.42 ask for the stand dev and variance. Do we have to show the spss output for those figures or is it okay to just write it down?
Re: 2.55 Do we really have to make 17 descriptives tables for each of the variables in 2.55? There are 17 variables and it says to make one for each variable.
Response: That does sounds rather redundant to do 17 descriptive tables. I would think a few representative tables should be sufficient. Thank you for bringing that to my attention.
RE: Does it matter how much work we show?
Response: You are not required to show the SPSS output (although it is useful when they ask for graphs and such). If the problem just calls for the stdev and variance you can just type (or write) them on your home work.
Post a Comment
<< Home