Upload
avram-caldwell
View
18
Download
2
Embed Size (px)
DESCRIPTION
The Research Process. First,. Collect data and make sure that everything is coded properly, things are not missing. Do this for whatever program your using (SPSS, EXCEL, etc.) If your using SPSS, its easy to check the data because you will run frequencies. - PowerPoint PPT Presentation
Citation preview
The Research Process
First,
• Collect data and make sure that everything is coded properly, things are not missing. Do this for whatever program your using (SPSS, EXCEL, etc.)
• If your using SPSS, its easy to check the data because you will run frequencies.
CHLDIDEL Ideal Number of Children
9 .6 .9 .9
36 2.4 3.7 4.7
553 36.9 57.3 62.0
205 13.7 21.2 83.2
87 5.8 9.0 92.2
11 .7 1.1 93.4
3 .2 .3 93.7
3 .2 .3 94.0
58 3.9 6.0 100.0
965 64.3 100.0
485 32.3
50 3.3
535 35.7
1500 100.0
0
1
2
3
4
5
6
7 Seven+
8 Aa Many as Want
Total
Valid
-1 NAP
9 DK,NA
Total
Missing
Total
Frequency Percent Valid PercentCumulative
Percent
If anything looks odd, you can fix it, was everyone asked the question?
SEX Respondent's Sex
641 42.7 42.7 42.7
859 57.3 57.3 100.0
1500 100.0 100.0
1 Male
2 Female
Total
ValidFrequency Percent Valid Percent
CumulativePercent
INCOME4 Total Family Income
585 39.0 39.0 39.0
300 20.0 20.0 59.0
230 15.3 15.3 74.3
385 25.7 25.7 100.0
1500 100.0 100.0
1.00 24,999 or less
2.00 25,000 to 39,999
3.00 40,000 to 59,999
4.00 60,000 or more
Total
ValidFrequency Percent Valid Percent
CumulativePercent
Next, run crosstabs and look for patterns..
SEX Respondent's Sex * INCOME4 Total Family Income Crosstabulation
224 143 118 156 641
38.3% 47.7% 51.3% 40.5% 42.7%
361 157 112 229 859
61.7% 52.3% 48.7% 59.5% 57.3%
585 300 230 385 1500
100.0% 100.0% 100.0% 100.0% 100.0%
Count
% within INCOME4 Total Family Income
Count
% within INCOME4 Total Family Income
Count
% within INCOME4 Total Family Income
1 Male
2 Female
SEX Respondent'sSex
Total
24,999 orless
25,000 to39,999
40,000 to59,999
60,000 ormore
INCOME4 Total Family Income
Total
Is the difference you see in sex and income significant or due to an error in the way you collected the sample?
• Determine the appropriate statistical tests to tell you this (refer to charts and examples in your text)
• Here we have sex which is nominal, but our dependent variable is on a scale. Why can’t you use correlation here?
Because…
When one thing goes up,
the other must go up (or down)
Sex can’t go up (at least not in this example)
In this instance, we use an independent samples t test to compare the means of both
group or ANOVA
Independent Samples Test
6.122 .013 1.224 1498 .221 .0781 .06380 -.04707 .20320
1.233 1414.912 .218 .0781 .06332 -.04614 .20227
Equal variancesassumed
Equal variancesnot assumed
INCOME4 TotalFamily Income
F Sig.
Levene's Test forEquality of Variances
t df Sig. (2-tailed)Mean
DifferenceStd. ErrorDifference Lower Upper
95% ConfidenceInterval of the
Difference
t-test for Equality of Means
Our results are not significant.
• We can’t show that any differences we see in income level are due to sex..but wait…could it be that we need to look at single males vs. single females?? Strange recode? Hmmm..the adventure continues..
Create the report