//  Task 1 routine setup
 
    capture log close 
    set more off 
    estimates clear 
   
    log using cda02c-ex2done.log, replace text
    version 9.2
    set scheme s2mono 
    
//  pgm:        cda02c-ex2done.do
//  task:       2 - Linear Regression - Excercise
//  project:    CDA Lab Guide
//  author:     your name
//  date:       today's date

    use science3, clear

//  Task 2: Keep four variables including cit9, fel, female and mcit3.
//  Examine these variables, drop missing cases, and draw scatter plots.

//  Step 2.1 keep the variables cit9, fel, female and mcit3.

    keep cit9 fel female mcit3

//  Step 2.2 examine variables
    
    codebook , compact
    sum cit9 fel mcit3
    tab female , m
       
//  Step 2.3 drop missing cases.
    
    mark nomiss
    markout nomiss cit9 fel female mcit3
    tab nomiss
    keep if nomiss==1
    drop nomiss

//  Step 2.4 graph matrix to explore your variables.

    graph matrix cit9 fel female mcit3, half
    graph export cda02c-ex2-fig1.emf, replace

//  Task 3: Estimate an OLS regression and list coefficients.

//  Step 3.1 run a regression of cit9 on fel, female and mcit3.

    regress cit9 fel female mcit3

//  Step 3.2 list the standardized coefficients

    listcoef, help

//  Task 4: Interpret the standardized and unstandardized coefficients.

//  Task 4a: Interpret the unstandardized effect of fel on cit9.

/*  Type your answer here

    For a unit increase in a scientist's doctoral prestige, the number
    of citations received by a scientists during years 6 to 9 are
    expected to increase by 3.7, holding all other variables constant.
*/

//  Task 4b: Interpret the standardized effects of fel on cit9.

/*  Type your answer here

    ** x-standardized:
    For a standard deviation increase in the prestige of a scientist's
    doctoral program, the number of citations is expected to increases
    by 3.6, holding all other variables constant.

    ** y-standardized:
    For a unit increase in doctoral prestige, a scientist's expected
    number of citations is expected to increase by 0.14 standard deviations,
    holding all other variables constant.

    ** fully-standardized:
    For a standard deviation increase in doctoral prestige, the number of
    citations a scientist receives in years 6 to 9 is expected to increase
    by .14 standard deviations, holding all other variables constant.
*/

//  Task 4c: Interpret the unstandardized effect of being female on cit9.

/*  Type your answer here

    Being a female decreases a scientist's citations in years six through
    nine by 8.2, holding all other variables constant.
*/

//  Step 4d Interpret the standardized effects of being female on cit9.

/*  Type your answer here

    ** y-standardized
    Being a female decreases the expected number of citations by 0.325
    standard deviation, holding all other variables constant.
*/

log close
exit