We invite you to shape the future of IBM, including product roadmaps, by submitting ideas that matter to you the most. Here's how it works:
Post your ideas
Post ideas and requests to enhance a product or service. Take a look at ideas others have posted and upvote them if they matter to you,
Post an idea
Upvote ideas that matter most to you
Get feedback from the IBM team to refine your idea
Help IBM prioritize your ideas and requests
The IBM team may need your help to refine the ideas so they may ask for more information or feedback. The product management team will then decide if they can begin working on your idea. If they can start during the next development cycle, they will put the idea on the priority list. Each team at IBM works on a different schedule, where some ideas can be implemented right away, others may be placed on a different schedule.
Receive notification on the decision
Some ideas can be implemented at IBM, while others may not fit within the development plans for the product. In either case, the team will let you know as soon as possible. In some cases, we may be able to find alternatives for ideas which cannot be implemented in a reasonable time.
Save the Random Effects of Mixed/Multilevel Models
Mixed/multilevel models are used to analyze hierarchical data (e.g., repeated-measures data, â€œnestedâ€ data such as when data is collected from individuals working in teams). Mixed models are able to disentangle variance that occurs among the lowest level of observations (e.g., differences between measurements repeatedly collected from the same person, differences between individuals working together in a team) from variance that occurs at higher levels of observation (e.g., differences between the people from whom measures were repeatedly collected, differences between the teams in which people are working). Mixed models achieve this by estimating a different intercept and/or slope(s) for each higher-level group. These per-group intercept and slope estimates are called â€œrandom effects.â€
I am requesting that SPSS provide a new keyword on the SAVE subcommand of the MIXED procedure that will save all the random effects of a multilevel model in a new dataset. Other statistical programs can easily save random effect estimates, and the SPSS MIXED procedure is estimating these random effects when running the MIXED procedure, anyway.
Currently, there is a way to derive one set of random estimates in SPSS from mixed models that have only one random effect (e.g., only a random intercept). To do this, you save the fixed predicted values and the random predicted values and use the difference between them to back-infer the random estimates for each group. However, this only works if you are modeling one random effect (e.g., one random slope with no random intercept). Most of the time, researchers who want to save random slopes are also estimating random intercepts to account for mean differences between their groups, so this hack is insufficient. Moreover, even when a model only estimates one random effect, it would be nice if researchers could simply save the random effect through the SAVE subcommand without needing to calculate it from the predicted values.
Please note that this request applies to the MIXED procedure on any operating system. The requested feature form required me to select one operating system when submitting this request, so I selected Windows because it is common among SPSS users.
Do not place IBM confidential, company confidential, or personal information into any field.