All Collections
Preparing and Loading Data
FAQ for Self-Loading .sav Files
FAQ for Self-Loading .sav Files

Tips to help with common questions related to self-loading data

Ian Brown avatar
Written by Ian Brown
Updated this week

KnowledgeHound's self-loading capability for .sav files is an easy way to quickly and easily load data for analysis. The article is intended to help with some common questions you may come across as you load data yourself. Many of these troubleshooting steps require making changes to the .sav file in SPSS.

For more information on self-loading, visit this article. For our general data quality guidelines, visit this article.

Question: Why isn't my data showing up at all?

Possible answers:

  • Self-loading requires a .sav or a file for analysis. Double-check to make sure your file format is correct.

  • If the file format is correct, the checkbox to "Use Professional Data Services?" should be available. Make sure to leave it unchecked as pictured below:

  • If you've successfully uploaded a dataset, you should see the three processing steps under the "Your data is processing" headline. If you notice progress has halted completely for an extended period (roughly 10 minutes), try refreshing your page. If nothing has changed, it may be due to the size of your dataset. Compressing your data to create a file, and uploading that instead may help speed the process up.

Question: Why are so many variables coming into KnowledgeHound as open-ends?

Possible answer: Try checking your variable types in SPSS to make sure they are not "strings".

Question: Why aren't my grid/matrix questions showing up how they look in SPSS?

First, it's important to understand what KnowledgeHound looks for in order to create a grid variable out of individual variables. There are two determining factors in "Grid Detection":

  1. Matching response values (except for numerics, which are all eligible to be combined as grids)

  2. Text patterns in variable labels

With that in mind, there are a few reasons why your grids may not be showing up as you would expect in KnowledgeHound.

  • Variables are labeled in a way that does not produce a pattern.

    • It is helpful to format grid question labels so that every label has an identical string of text

  • If one of the series of the grid has no response data, it may not be included by default. When you publish the dataset, the variables with no response data will disappear from the variable list entirely.

    • Variables may also disappear after publishing for other reasons outside of having no response data. See the next question, "Why are some of my variables missing?" for more details.

Question: Why are some of my variables missing?

Certain variables may be omitted in KnowledgeHound due to rules in the system. A few common examples of those rules are as follows:

A variable will be omitted if…

  • The variable has no response data

  • The same response was chosen by all respondents

  • The variable label contains the word "hidden"

  • The variable is unlabeled

Question: Where is my weight variable?

Weight variables in your dataset should be available to be used for weighting in analysis on KnowledgeHound. If you have a weight variable in your data that isn't showing up in KnowledgeHound, here is a possible reason why:

  • The weight variable does not have a value assigned for every row. Besides being a numeric variable, KnowledgeHound looks for weight variables that have values attributed to every respondent so that no rows have a null value. If your weight variable does not apply to the entire sample of your study, you may have to fill in null rows with 0 or 1 for it to be recognized as a weight variable.

For more information on working with weight variables in KnowledgeHound, visit this article.

Question: How do I set a default weight variable for my study?

Answer: Today, KnowledgeHound must manually set your default weight for you if you would like it to change. Reach out to KnowledgeHound using the blue chat bubble on the bottom-right corner of your screen while using KnowledgeHound.

Did this answer your question?