In descriptive statistics, we analyze the sample data and then we try to draw conclusion about the sample data only, without trying to generalize the conclusions to the whole population. In this sense, we are simply "describing" our sample, hence the name descriptive statistics.
To research is to answer some predefined questions systematically, using data and logical arguments, and with reference to previous works. To do statistics (统计) is to collect (统), summarize (计), and draw conclusions from the data so as to answer the research questions. Therefore:
In data analysis, we DO NOT think in terms of how many or what types of charts and tables have been generated. Instead, we ask ourselves: Have I answered ALL the research questions that could be answered by the data? If yes, then we are done. Otherwise, we need to keep on extracting more information from the data.
Data -> Validation -> Validate data -> Single-Variable Rules (This is for checking the distributions at a glance.) (But not available in some setups of SPSS.)
Toolbar button -> Run descriptive statistics
Analyze -> Descriptive Statistics -> Frequencies (This one is the most useful)
Click the Statistics button to choose percentiles, central tendency, dispersion, and distribution options
Click the Charts button to choose histograms if needed. E.g., You can plot the histogram with a normal curve to check if the distribution is normal or near normal. (But do not choose bar chart or pie chart, since they are not applicable to scale level data.)
Double click on the output to customize display settings
Paste the tables / charts to Microsoft Word
Analyze -> Descriptive Statistics -> Descriptives
Click the Options button to choose mean, sum, and other statistics under dispersion, and distribution if needed
Analyze -> Descriptive Statistics -> Explore
Click the Statistics button to choose the statistics
Click the Plots button to choose box plots, stem-and-leaf, histogram, or normality plots with test.
Analyze -> Descriptive Statistics -> Frequencies
Select the "Display frequency tables" option in the lower left hand corner of the dialog if needed.
Click the Statistics button to choose the required statistics (but note that many of these statistics are not applicable to categorical data).
Click the Charts button to choose bar chart or pie chart if needed.
Double click on the output to customize display settings
Paste the tables / charts to Microsoft Word
Analyze -> Descriptive Statistics -> Crosstabs
Put the independent variable in the Row and dependent variable in the Column.
Click Statistics button and uncheck all options
Click Cells button and optionally check Observed and Expected counts, as well as percentages. Uncheck all others.
Inspect the cross-tabulation in the output.
Consider the meaning of the Expected Count.
Graphs -> Chart Builder
Play with the different types of graphs, esp. histogram, box plot
Try clustering and panels to split the charts by variables
Analyses->Exploration->Descriptives
Select the variables to be analyzed.
Add grouping variables (if any) into the “split by” field.
Select the appropriate statistics or plots you want to add.
Analyses->Exploration->Descriptives (Same as above, but look at the frequency table and charts only)
Analyses->Frequencies->Independent Samples:
Put the independent variable in the Rows and dependent variable in the Columns.
Click Cells and optionally check Observed and Expected counts, as well as percentages.
Inspect the cross-tabulation (contingency tables) in the output.
Jamovi is not yet very good at drawing graphs. Limited graphing options can be found in the options under the corresponding procedures.
It depends on the levels of measurement of your data.
Scale data:
Descriptive statistics: Mean, standard deviation, histogram, correlation, regression, etc.
Inferential statistics: t-test, ANOVA, etc.
Ordinal data (e.g. Likert scale):
Descriptive statistics: Median, frequency table, bar chart, etc.
Inferential statistics: non-parametric tests
For scale and ordinal data, you can either analyze all the cases as a whole, or (most often) you can also do some cross-comparisons. e.g.,
Independent samples: Compare your data across demographic grouping variables (e.g. gender, income range, academic level).
Paired / dependent samples: Compare your data across two different items obtained from the same respondents (e.g. pre-test vs. post-test).
Nominal data:
Descriptive statistics: frequency table, cross-tabulation, bar chart, etc.
Inferential statistics: chi-square test, etc.
Qualitative data (e.g. open-ended comments):
Thematic analysis, discourse analysis, etc.
In SPSS, use Data -> Split File to split the variables by grouping variables. All the analysis from this point on will be conducted separately to each group.
In Jamovi, assign “split-by” variables or grouping variables in the corresponding analyses procedures.
Likert scale is ordinal level. You can analyze its frequency distribution, draw bar charts, find the median, or conduct non-parametric tests. Just remember to avoid mean, standard deviation, t-test, ANOVA, etc. or other analyses specifically for scale level data.
You can rename your output in the left pane of your output window. Also remove the outputs that you do not want.
In SPSS, you can use the Graph Builder for scatter plot.
In Jamovi, you can do this with the scatr module.
In SPSS, if you have variables X1, X2, X3 and you want to find their mean, you can click Transform -> Compute Variable and then input the formula MEAN(X1,X2,X3) in the Numeric Expression box. See here for details: https://libguides.library.kent.edu/spss/computevariables.
In Jamovi, click the column header of the last variable you want to include in the calculation (e.g., A4), and then click Data->Compute:
Name the new variable and add a description if needed.
Type the formula for the compute, e.g., =MEAN(X1, X2, X3). You can click the fx button for some available functions and the names of the variables.
Note: You can do other calculations on the variables in a similar manner. e.g. To find the sum between X1 and X2 you can input X1+X2 instead.
There are a few ways to check if your data follow normal distribution:
Method 1: Plot the histogram of your variable and inspect it manually
SPSS: Analyze -> Descriptive Statistics -> Frequencies -> Put the variables you want to test into the Variable(s) box -> Click the Charts button -> Choose Histograms -> Also check "Show normal curve on histogram". You can then inspect the resulting histogram manually.
Jamovi: Analyses -> Descriptives -> Plots -> Histograms -> Check Histogram
Method 2: Do a Q-Q plot of the data
SPSS: Analyze -> Descriptive Statistics -> Q-Q Plots -> Choose the variables to test -> Choose Normal in Test Distribution -> OK
Jamovi: Analyses -> Descriptives -> Plots -> Q-Q Plots -> Check Q-Q
If the resulting data points in the Q-Q plot fit well into the straight line, then the data is normally distributed.
Method 3: Do the Shapiro-Wilk test
SPSS: Analyze -> Descriptive Statistics -> Explore -> Put the variables into the Dependent List -> Click Plots -> Check Normality plots with tests. This will give you the Shapiro-Wilk test (and the Kolmogoov-Smirnov test as well)
Jamovi: Analyses -> Descriptives -> Statistics -> Normality -> Check Shapiro-Wilk
If p<0.05 for these tests, it means your data is not normally distributed.
But note that these tests are quite sensitive - they will tell you that your data is not normally distributed even if it is only slightly deviated from that. I would usually recommend the first two methods instead.
Finally, in Jamovi you may also find these normality test options in the individual hypothesis test procedures. For example, in One-Way ANOVA in Jamovi, you can find Normality test and Q-Q Plot under Assumption Checks.