Assignment 1: Jamovi Introduction
Using the NELS dataset
This assignment uses the NELS dataset as an example. Download the dataset before beginning.
Getting started in Jamovi
- Open Jamovi.
- Select menu (three lines) in the upper-left corner, choose Open, and select This PC. Click Browse to locate the downloaded NELS data file. You can also drag the file into the Jamovi window.
- Jamovi displays the data in a spreadsheet. Each row is a student and each column is a variable.
- Closing Jamovi: my recommendation for this class is to NOT save anything. When you get a prompt to “save XXX?”, you should always choose “No.” If you need to re-access statistical results, you can simply re-run the analysis. Or, you can always take screenshots as needed.
- (Optional) If you really do want to save your work (e.g., for a future project), use the menu button and choose Save As. Jamovi saves the dataset and analyses together in a
.omvfile.
- (Optional) If you really do want to save your work (e.g., for a future project), use the menu button and choose Save As. Jamovi saves the dataset and analyses together in a
Data and variable views
- Select the Data tab (at the top) to view the dataset and inspect variables.
- Under data view, each column corresponds to a variable. To see a variable’s settings, select the column and choose Setup. You can also double-click a column header. Note that the setup panel shows the variable name, description, data type, and measure type. The measure types are Nominal, Ordinal, Continuous, and ID.
- For this class: don’t worry too much about the “ID” measure type. Also, don’t worry about the “data type” (integer, decimal, text).
Jamovi analysis examples
These tools are introduced in Chapter 2 and will be used throughout the semester.
Frequency tables (Nominal or Ordinal)
- Select Analyses > Exploration > Descriptives.
- Move
regioninto the Variables box. - Under Statistics, select Frequency tables. You can add several variables to the same analysis to create multiple tables.
- Note that this does NOT work for continuous variables.
Percentile and percentile rank
- Open Analyses > Exploration > Descriptives.
- Move
achmat08into the Variables box. - Under Statistics > Percentile Values, select
cut points for 10 equal groups(you need to enter10). - Jamovi only allows up to 10 equally spaced cut points, so you cannot get precise percentile or percentile rank values for continuous variables beyond these cut points. With
achmat08:- Example 1: What is the 28th percentile? Answer: between 47.8 and 51.
- Example 2: What is the percentile rank of 50? Answer: between 20% and 30%.
Statistic Value N 500 Missing 0 Mean 56.6 Median 56.2 Standard deviation 9.34 Minimum 36.6 Maximum 77.2 10th percentile 44.2 20th percentile 47.8 30th percentile 51.0 40th percentile 53.7 50th percentile 56.2 60th percentile 58.9 70th percentile 62.1 80th percentile 65.7 90th percentile 68.9 100th percentile 77.2
Summary statistics and plots (univariate)
- Select Analyses > Exploration > Descriptives.
- Move
achmat08into the Variables box. - Under Statistics, select the desired numerical summaries (e.g., mean, median, standard deviation). Under Plots, select the desired visual summaries (e.g., histogram, box plot, bar plot).
Practice Questions
Use the NELS dataset in Jamovi. The same information can often be generated with more than one Jamovi option.
- How many people are recorded in the NELS dataset? How many are women and how many are men (
gender)? - How many variables are recorded in total?
- How many regions are there? What are the regions?
- What does the variable
parmarl8represent? If a person has a value of 1 for this variable, what does it mean? - Find the person whose ID is 10. What is the highest level of education this person expects to get?
- What percentage of students in the dataset have ever smoked marijuana (
marijuan)? - What is the biggest family size recorded?
- What is the recorded measure type for
alcbinge? What is the actual appropriate measure type for this variable?
Show solutions
- There are 500 people recorded in the NELS dataset: 227 are male and 273 are female.
- 47 variables are recorded in total.
- There are 4 regions: Northeast, North Central, South, and West.
parmarl8represents a person’s parents’ marital status while they were in eighth grade. A value of 1 means that the person’s parents were divorced.- The person expects to earn a Ph.D., MD, JD, or another professional degree.
- 18.4% of students in the dataset have ever smoked marijuana.
- The biggest family size recorded is 9.
- The recorded measure type for
alcbingeis nominal, but the appropriate measure type is dichotomous.