1 Unit 1: Introduction to Univariate & Bivariate Analysis using SPSS The three parts of SPSS: the data manager, data files, and output files Loading data files, changing data sets, using the utilities function to examine variable values Univariate analysis o Frequencies o Valid Percent o Cumulative Percent Examining a bivariate relationship between dichotomous variables o Crosstabulations Reading 2 x 2 tables Data loss Cells Column, row, and total percentages o Statistics Pearson Chi Square Cramer’s V Reference information Pearson Chi Square p. = probability that results occurred by chance <value= <likely by chance .05 or less = accepted as significant (not chance occurrence) Cramer’s V 0- 1.00 = strength of association <.10 = weak .10-.30 = moderate >.30 = strong Sample Table: Crosstabulation of support for capital punishment by sex , with column percentages reported. Respondent’s Sex total male female 82.6% 71.6% 76.6% Capital favor Punishment 17.4% 28.4% 23.4% oppose total 100.0% 100.0% 100.0% Pearson X2:p.<.000; Cramer’s V=.129 Source: 1996 GSS Tab subset Sample results statement for bivariate crosstabulation of nominal level dichotomous variables: While a majority (76.6%) of respondents in the 1996 GSS sample support capital punishment (DPS), there is a moderately strong (V= .129), statistically significant difference (p.<.000) in DPS by sex, with 82.6% of men favoring capital punishment, in contrast to 71.6% of women. 2 Univariate Analysis: Examining a single variable Ex. 1) Fill in the frequencies and valid percents for the following variables. Data set Variable Attributes GSS96TAB CAPPUN Favor 1 Oppose 2 Liberal 1 Moderate 2 conservative 3 Extremely liberal 1 Liberal 2 Slightly liberal 3 Moderate 4 Slightly conservative 5 Conservative 6 Very conservative 7 Favor 1 Oppose 2 GSS96TAB GSS96COR GSS72SAV POLVIEWS POLVIEWS CAPPUN2 Values GSS96TAB DEGREE At least some college 2 GSS72.SSAV DEGREE At least some college 2 GSS96COR Percent of respondents who identify themselves as liberal to some degree? Frequency Valid % 3 Bivariate Analysis: Examining the relationship between two variables. Ex. 2) Does support for capital punishment vary by sex in the 1996 GSS subset? o o o o o Analyze----Descriptives----Crosstabs Row: Dependent variable (D.V.) (Cappun) Column: Independent variable (I.V.) (Sex) Cells: Column & total Statistics: Chi Square (X2) & Phi/Cramer’s V Is the difference statistically significant? (low chance of being random) USE: Pearson Chi Square: p. = probability that results occurred by chance <value= <likely by chance .05 or less = accepted as significant If it is significant, how strong is the association? USE: Cramer’s V 0- 1.00 = strength of association <.10 = weak .10-.30 = moderate >.30 = strong Fill in the following table and sentences with information from the SPSS crosstabulation of SEX & CAPPUN in the GSS96TAB data set. If the answer to a question is not included in the chart, write: N/A. (Indicate N/A even if you could calculate the answer, but it is not directly observable). Table 1: Crosstabulation of support for capital punishment by sex , with column percentages reported. Respondent’s Sex total male female favor Capital Punishment oppose total ; Cramer’s V= Pearson X2: p Source: 1996 GSS Tab subset The total number of respondents (n) with data used in the crosstabulation = ________ The number of females who oppose the death penalty = __________ The percent of males who favor the death penalty = ____________ The percent of those favoring the death penalty who are female = __________ The total number of females responding to the death penalty question = __________ The percent of respondents opposing the death penalty = _____________ The percent of those opposing the death penalty who are male= ______________ Are the sex differences in support for capital punishment statistically significant? _________ If so, how strong are the sex differences? _____________ (SEE COVER PAGE FOR A WRITTEN DESCRIPTION OF THE DATA IN THIS TABLE) 4 Ex. 3) Practice Question: Do men and women differ in fear in the 1996 GSS subset? (GSS FEAR question = Is there any area right around here-that is, within a mile-where you would be afraid to walk alone at night? Yes- 1 No-2 Table 1: Crosstabulation of support for capital punishment by sex , with column percentages reported. Respondent’s Sex male female total yes Fear no total Pearson X2: p ; Cramer’s V= Source: 1996 GSS Tab subset The total number of respondents (n) with data used in the crosstabulation = ________ The number of females who are afraid to walk alone at night = __________ The percent of males who are not afraid to walk alone at night = ____________ The percent of those who are afraid to walk alone at night who are female = __________ The total number of females responding to the fear question = __________ The percent of respondents who are afraid to walk alone at night = _____________ The percent of those who are not afraid who are male= ______________ Are the sex differences in fear statistically significant? _________ If so, how strong are the sex differences? _____________ While a majority ( ) of respondents in the 1996 GSS sample say that they are/ are not afraid to walk alone at night, there is a _____________ (V. ), statistically significant difference (p. with _________of men expressing fear, in contrast to __________of women. ) in fear by sex, 5 Unit 1: Lab Review exercise: Bivariate analysis of nominal level dichotomous variables. Fill in the following chart. Enter N/A under Cramer’s V if P is not significant. Sample Table: Crosstabulation of support for capital punishment by sex , with column percentages reported. Respondent’s Sex total male female 82.6% 71.6% 76.6% Capital favor Punishment 17.4% 28.4% 23.4% oppose total 100.0% 100.0% 100.0% Pearson X2:p.<.000; Cramer’s V=.129 Source: 1996 GSS Tab subset DATA SET DEP.VAR. IND. VAR. GSS96TAB EXAMPLE CAPPUN SEX GSS96TAB SPANKING RACE GSS96TAB TRUST SEX GSS96TAB GUNLAW CAPPUN GSS96TAB FEWORK SEX GSS96TAB VOTE92 RACE GSS72SAV FEWORK SEX GSS72SAV VOTE68 RACE GSS72SAV CAPPUN SEX CHI SQ P P<.000 SIG? YES/ NO yes CRAMER’S V .129 STRENGTH I.V/DV ATT1% moderate Male 82.6% favor I.V/DV ATT2 % Female 71.6% favor 6 Answer the following questions referring to lab review exercise above and using the results reporting paragraph demonstrated in this packet. Q.1) Are race differences in support for spanking evident in the 1996 GSS data set? Q.2) How have racial differences in voting changed from the 1972 GSS sample to the 1996 GSS sample and how might you explain this sociologically? Q.3) How have sex differences in attitudes toward women working changed from the 1972 GSS sample to the 1996 GSS sample and how might you explain this sociologically?