# Thread: Influence of a person in repeat measurement (multilevel) or longitudinal data

1. ## Influence of a person in repeat measurement (multilevel) or longitudinal data

Hey,

I have a dataset of 50 people with 4 responses over the year. Responses were a 5-point Likert scale collapsed into a binary variable (0,1). These 50 people were examined on 10 items, so 4 responses per each of the 10 items. I had previously ran 10 logistic multilevel analyses to show a positive trend in responses toward having a 1 over the year within each of the items.

I am fine running these analyses and I am happy with the results. Though, now I am thinking some people may be more likely to be an "0" across multiple items across time. It may be apart of their personality or demeanor. I was wondering if anyone had ideas or input on how to examine whether a person tended to be classified as an "0" in multiple items. So for an example say the item was confidence riding a bike, then another was confidence driving a car, etc. But the hitch is that I have 4 time points per person for 10 items.

Perhaps an initial approach may be finding the average response (0-1) for each person and see if any individual jumps out as having a low average.

2. ## Re: Influence of a person in repeat measurement (multilevel) or longitudinal data

Within the same idea, does anyone know of a theory of self-confidence? So you ask a person about their confidence, and even though they may be good at the task some people tend to report lower self-confidence than others of equal skillset.

3. ## Re: Influence of a person in repeat measurement (multilevel) or longitudinal data

Originally Posted by hlsmith
Within the same idea, does anyone know of a theory of self-confidence? So you ask a person about their confidence, and even though they may be good at the task some people tend to report lower self-confidence than others of equal skillset.
I am not familiar with that specific theory, but would say that it is more complex than how you describe the theory. For example, a person may be highly confident, but through cultural, family or religious upbringing will refrain from self aggrandizement or even the appearance of such.

I think you would have to observe actions under circumstances that would elicit the behavior. There are countless examples in military situations where the "self confident" panic under fire, and the quiet, understated person shows extreme heroism. I have seen similar, though less dramatic examples in a work environment.

4. ## Re: Influence of a person in repeat measurement (multilevel) or longitudinal data

True.

My colleague who works with this sample, stated one of the "Not Confident" individuals is actually very good at the tasks. So, if I could go back, I would have had independent external raters as well to address this phenomena.

 Tweet

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts