230 likes | 416 Views
Session 3. Tables in Stata. Importance of tables. The analysis of many surveys is primarily Through giving tables And then interpreting the information Simple tables are one-way frequencies Or of percentages Or Multi-way frequencies (or percentages)
E N D
Session 3 Tables in Stata
Importance of tables • The analysis of many surveys is primarily • Through giving tables • And then interpreting the information • Simple tables are one-way frequencies • Or of percentages • Or Multi-way frequencies (or percentages) • Tables can also contain summary statistics • Like means
Objectives • Particiants should be able to • Produce tables of each defined type using Stata • Interpret data from these tables • Produce tables of interest for the data set created for the district in the previous session.
Menus for Stata tables To explore table facilities, we will open the data file K_combined_labelled.dta
Different menu options Match these options to the different types of table that are needed to process survey data
One-way table Open the data file K_combined_labelled.dta Statistics => Summaries, tables & tests => Tables => One-way tables
Results for a one-way table Check you can interpret these results
Multiple one-way tables Statistics => Summaries, tables & tests => Tables => Multiple one-way tables
Tables using commands • The commands are simple. • For one-way tables try • tab q31 • For several one-way tables, use tab1, e.g. • tab1 region rurban q11 a31 • For two-way tables, try • tab q11 q16
Try the options Use the dialogue – with the options for Missing and/or sorting • Type tab1 q126 q127 q128 q129 q31, missing sort • What effects do the options sort and missing have?
Your turnOne-way and two-way tables • Work through sections 7.1.4 and 7.2 in the Guide. • Include 7.2.1 where you check for coding • Include 7.2.2 for lists of two-way tables • Type tab q31 q126, • Then repeat with tab replaced by tab1. • Explain the difference between tab1 and tab to your neighbour.
Percentages – interpret results Interpet the results from this table Repeat, but getting column percentages. Interpret those results Hence explain which set of percentages is more appropriate in this table
Multi-way tables • Can use tabulate command • with by prefix. For example: • bysort rurban: tab q128 q11, row • Or the “two-way tables” dialogue • Or use the more powerful table command • table q128 q11, contents(freq) by(rurban) row col • Or the “Tables of summary statistics (tables) dialogue • See result in next slide
A three-way table Interpret these results Then use the Two-way tables (with By) To give row or column percentages
Second example Use the dialogue or the command table q11 rurban q126, by(q16) Where the row var = Sex (q11) Column variable = Rurban Supercolumn variable = Employed (q126) Superrow variable =Literacy (q16) Interpret the table Try the variables in different orders, to see if your interpretation is simplified
Your turn • If you need more time on multi-way tables • Then go through Section 7.3 in the guide • Then go through Sections 7.4 and 7.5
Tables of summary statistics • Section 7.4 and 7.5 consider summary statistics • For simple tables they use the tabstat dialogue (or command) • Or the table dialogue (or command) is used
Dialogue for tabstat command Menu sequence: Statistics => Summaries, tables & tests => Tables =>Table of summary statistics (tabstat) (with Option: Use as columns Statistics)
Summary statistics in a table • Note: • qd41, qd42, qd43 are expenditures in the previous month for purchase of maize, bean and other seed respectively • Omitting the option col(stat) would give the statistics as rows rather than as columns as shown above
Dialogue for table command Menu sequence: Statistics => Summaries, tables & tests => Tables => Table of summary statistics
Summary statistics in a table Discuss with your neighbour the difference between use of tabstat and table .
Further practical work • Load your own data into Stata, and produce tables that address an objective of interest that you wish to explore with your data.