Dear Rachel,

I think Federico’s advice is on the point. Truth table rows are never deleted in QCA. Sometimes, it looks like they were deleted, but they are always present in the background. Literally deleting rows confuses the software because the truth table becomes incomplete. Rows with cases and consistency less than 0.8 should be coded zero. Rows with no cases should be treated as logical remainders, which is not the same as coding them 0. When using the delete-and-code option inside the truth-table-algorithm window, these are the rows that disappear, but are still present in the background.

Regards, Ingo

Hello,

I am attempting to teach myself how to use the fsQCA 2.5 software and I continuously receive error messages when I get to the “analysis” stage.

Below I describe the situation, and I appreciate any help this forum is able to provide. Thank you in advance!

I imported my data without any trouble. It consists of 31 Cases/Rows, each a different teen pregnancy prevention program, with about 20 Columns/Variables that represent attributes that are coded present (1) or absent (0). For example, one Column/Variable of interest is “School” and is coded =1 if the pregnancy prevention program is delivered in a school-based setting, and it is coded =0 if it is delivered in a non-school setting, such as a clinic.

Our research question for which we thought crisp set QCA would be useful is the following: what configurations of attributes are associated with “successful” programs. The binary outcome “successful” (1/0) is considered present if the teen pregnancy prevention program has a significant impact (evaluated by RCT) on any  of certain pregnancy prevention behaviors (e.g. delayed age of initiation of sex or reported using condoms for most recent sexual encounter).

So far, I have been able to move through the following steps:

Analyze à Crisp Truth Table Algorithm à Outcome = anyoutcome (1 = program was successful on any of several desired behavioral outcomes, 0 = program was not successful on any of the desired behavioral outcomes) à Causal Conditions = school (school-based program), 10-13 year olds (program targeted pre-adolescents ages 10-13 years), 14-15 year olds (program targeted young high school age adolescents, ages 14-15 years), 16-18 year olds (program targeted older adolescents), psychosocial skill building (program teachers psychosocial skills), behavioral skill building (program teachers behavioral skills), etc.

When I select a set of causal conditions, the program generates a Truth Table. I then manually code the outcome =1 for all those with consistency scores of .80 or greater, I manually code the outcome =0 for all those configurations with consistencies less than or equal to .30, and I delete the rows with no actual configurations existing or with consistencies between .30 and .80.

When I click on Standard Analysis, I get error messages such as: “ERROR (Quine-McCluskey): The 1 Matrix is Empty.”

Followed by: “Error: can’t read “rows”: no such variable.”

If I try Specify Analysis, I get an Intermediate Solution with no information, such as the following:

--- INTERMEDIATE SOLUTION ---

frequency cutoff: 1.000000

consistency cutoff: 1.000000

Assumptions:

Any help you can provide is most appreciated – thank you!

Rachel Gooze

Rachel A. Gooze, PhD, MPH |Research Scientist

7315 Wisconsin Ave, Ste 1200W | Bethesda, MD 20814

Office: (240) 223-9354

Email: [log in to unmask]">[log in to unmask] 

Research to Improve Children's Lives