#Stata summation how to#
#Stata summation code#

In the present case, this does not occur when any of the variables are removed. We are looking for any items which, if removed, would result in an Alpha that is appreciably higher than the one computed for the whole scale. This tends often to present an inverse picture to the corrected item-total correlation. In the rightmost column вЂњalpha,вЂќ we see the effect on Alpha if we excluded each of the items in turn. Ideally, all items will have moderate or strong correlations to the corrected item total as is the case for the six we have chosen. It can be used to understand how well each item вЂњfitsвЂќ with the others and the resultant scale. This shows the correlation between each of the items with a summative scale containing all of the other items. The first is the one that shows the corrected item-total correlations, labelled вЂњitem-rest correlationвЂќ. There are quite a few columns, but we only need focus on two of them. This output can be used to diagnose problems with particular questions. The police in this area are dealing with the things that matter to this communityДЇigure 5: Reliability Statistics in Stata.ДЇigure 6 shows a table of item-total statistics. The police in this area understand the issues that affect this community The police in this area can be relied on to deal with minor crimes The police in this area treat everyone fairly regardless of who they are The police in this area would treat you with respect if you had contact with them The police in this area can be relied on to be there when you need them These variables are in the form of Likert scales, where respondents select answers on a 5-point scale from agree strongly to disagree strongly to the following statements: Six items in the BCS tap these concepts amongst a sample of the adult population of Great Britain. Confidence in the police can be considered a multifaceted concept that includes perceptions of competence, effectiveness, fairness, procedural justice, and shared values. Since 2011, the survey has been renamed the Crime Survey for England and Wales. This survey is fielded each year to a probability sample of adults in Britain.

#Stata summation how to#

This example shows how to construct a scale that measures confidence in the police from a set of questions included in the 2007вЂ“8 British Crime Survey (BCS). If you really have a table with blanks that you have to assume take the value of previous obs, e.g., in the table image you sent, row 2 of column 'Variable' really should have the value 'Ice Cream Flavor 1', then first you've got to fix that so that the values of that variable are explicit instead of implicit (i.e., it should have 'Ice Cream Flavor 1', instead of being blank).ДЄnyway, once that's done, and if the unit of analysis is person, then after multiplying as suggested above, collapse can calculate the sum by flavor - or, if there are few enough combinations, you could use table with the ,contents(sum varname) option.2 An Example in Stata: Confidence in the Police As a matter of math, in line 2 you could instead generate t123 = t12 * topping3, but then you've got to remember how t12 was created, and those sorts of dependencies are murder for maintenance later in soooo many instances that it's best to avoid them.ДЇor the rest of the details, it depends a lot on the structure of the actual data. Generate t234 = topping2 * topping3 *topping4Д®tc. Generate t1234 = topping1 * topping2 * topping3 * topping4 Generate t123 = topping1 * topping2 * topping3

#Stata summation code#

I wouldn't bother with a loop there, because the number of lines of code is small and the loop logic would be more complex than just writing them: generate t12 = topping1 * topping2 Carry that through the combinations you need, and you'll have a set of indicators for each combination. topping1 * topping2 will equal 1 when both toppings are 1 and zero else. That said, I'd use the magic of multiplication here. Showing me an image of the table isn't quite as useful. Pop that data into Stata, then use dataex to get output we can use to help. There has to be an easier way!!! Is there a way to loop this? The problem is: there are 6 million flavors. My question is: how do I use STATA to calculate how many people who picked a particular ice cream flavor took any 2 toppings? any 3 toppings? any 4?Д¬ount if topping1=1 & topping2=1 & Flavor1=1Д¬ount if topping1=1 & topping3=1 & Flavor1=1Д¬ount if topping1=1 & topping4=1 & Flavor1=1Д¬ount if topping2=1 & topping3=1 & Flavor1=1Д¬ount if topping2=1 & topping4=1 & Flavor1=1Д¬ount if topping3=1 & topping4=1 & Flavor1=1ДЄnd then sum up the counts to get the per flavor total. Note that both Flavor and Toppings is a categorical variable: yes(1) or no(0) for each person. dataex Ice_cream Topping1 Topping2 Topping3 Topping 4 I've noticed you've commented on my previous posts (as well as the posts of others) with very helpful advice.