SPSS Statistics

 View Only
Expand all | Collapse all

Creating growth rate for a variety of variables

  • 1.  Creating growth rate for a variety of variables

    Posted Wed December 15, 2021 08:43 AM
    Hello, I am looking at ACS data to see how several variables (population, visitation, migration rate, rate of seasonal residence etc) impact GINI index in two sets of communities.

    I am planning to do an ANOVA with the variables listed above as the independent variables and GINI index as independent. However, I would like to use the growth rate for each year for this analysis. 

    How do you compute growth rate version of a variable you currently have? For each town I have information on all of the variables from 2010 to 2019. It is easy to average the growth rate in excel (mean of community group 1 population 2011- mean of community group 1 population 2010/ mean of community group 1 population 2010). But I need to find a way to do this SPSS.

    Using population as an example (Here I just included 2 towns but in reality I have 3.901 towns):

    Community Name             Year        Population       GINI                                          My question: Population yearly growth rate        GINI yearly growth rate
    Smithtown                         2010        21,000               .46
    Smithtown                         2011        22,551               .47
    Smithtown                         2012        20,543              .46   etc continuing to 2019
    Joetown                            2010          5,000              .45
    Joetown                            2011           4,900             .47  etc continuing to 2019

    ------------------------------
    Elizabeth Depew
    ------------------------------

    #SPSSStatistics


  • 2.  RE: Creating growth rate for a variety of variables

    Posted Wed December 15, 2021 11:13 AM
    When the data are ordered so that each year is it's own record in the dataset...
    DATA LIST FREE /CommunityName (A32) Year (F4)  Population (F8)  GINI (F4.2).
BEGIN DATA.
Smithtown     2010        21000   .46
Smithtown     2011        22551   .47
Smithtown     2012        20543   .46
Joetown       2010         5000   .45
Joetown       2011         4900   .47
END DATA.

    you can compute such a "change" variable like this:

    SORT CASES BY CommunityName.
DO IF CommunityName=LAG(CommunityName).
COMPUTE growth_rate=(100*(population-lag(population)) / lag(population)).
END IF.

*In this small dataset, you can verify the numbers are what you expect by doing a simple LIST command...

LIST.​

    Just one thing: It looks to me as if these data lack the assumption of independence between cases, which is necessary for ANOVA. This looks like more of a "nested" type analysis. I am by no means as expert in such cases but I think you should consider using GENLINMIXED with Towns as "subjects".

    ------------------------------
    Rick Marcantonio
    Quality Assurance
    IBM
    ------------------------------

    Original Message