'Dementia patient-caregiver connection, pronoun usage, July 2020' (AsPredicted #43985)
Author(s) This pre-registration is currently anonymous to enable blind peer-review. It has 5 authors.
Pre-registered on 2020/07/02 - 12:34 PM (PT)
1) Have any data been collected for this study already? It's complicated. We have already collected some data but explain in Question 8 why readers may consider this a valid pre-registration nevertheless.
2) What's the main question being asked or hypothesis being tested in this study? We are evaluating the relationship between pronoun usage and health and well-being outcomes in caregivers of patients with dementia.
Greater use of “we” pronouns in response to a question about caregiver connectedness to the patient will be associated with less concurrent and future depression, burden, and loneliness.
Greater use of 1st-person (e.g., "I" pronouns) and 3rd-person (e.g., "he"/"she"/"they") pronouns in response to a question about caregiver connectedness to the patient will be associated with more concurrent and future depression, burden, and loneliness.
3) Describe the key dependent variable(s) specifying how they will be measured. The dependent variables were caregiver depression, caregiver burden, and caregiver loneliness.
Caregiver depression was measured using the PHQ-9; caregiver burden was measured with the Zarit-12; caregiver loneliness was measured with a 3-item loneliness scale.
These three measures were collected via verbal self-report through a telephone interview at baseline and 6 months, 12 months, 24 months, 36 months, 42 months (after baseline).
4) How many and which conditions will participants be assigned to? Data will be drawn from an existing, longitudinal intervention study testing a model of dementia care and support. Participants were randomly assigned to one of two groups: a treatment condition or a standard-of-care condition. We will account for intervention group assignment in analyses, but this is not the primary focus of the present pre-registration.
5) Specify exactly which analyses you will conduct to examine the main question/hypothesis. We will conduct linear regression analyses with pronoun variables predicting caregiver outcomes (i.e., depression, burden, loneliness). First, we will conduct separate regressions with each type of pronoun (i.e., we-words, I-words, 3rd person-words) predicting outcomes. Next, we will conduct follow-up analyses including all pronoun types that were significant in the original regressions, predicting caregiver outcomes. We will evaluate the pronoun variables as both sums (i.e., total number of a pronoun type) and ratios (i.e., total number of a pronoun type divided by (a) total pronouns used, or (b) total words spoken).
6) Describe exactly how outliers will be defined and handled, and your precise rule(s) for excluding observations. Outliers will be defined by any values that are 3 standard deviations above or below the mean. We will exclude these values from the analyses.
7) How many observations will be collected or what will determine sample size? No need to justify decision, but be precise about exactly how the number will be determined. The full sample of caregivers have already been enrolled and completed this study. We will use all available data to maximize the sample size. However, we will exclude any caregiver responses that were incorrectly transcribed (i.e., the research coordinator did not record responses verbatim).
8) Anything else you would like to pre-register? (e.g., secondary analyses, variables collected for exploratory purposes, unusual analyses planned?) We will conduct the following secondary analyses:
(1) Latent growth curve models (LGMs) of caregiver depression, burden, and loneliness (respectively) in a SEM framework. We will include regressions with pronoun variables (i.e., we-words, I-words, 3rd person-words) predicting the latent slopes.
(2) Mediation analyses with intervention group (assigned at baseline) predicting future caregiver depression, burden, and loneliness, with pronoun variables measured at month 6 mediating these associations.
We have also collected a number of variables for exploratory purposes (see list below).
Patient: demographic info (race, age, gender, preferred language, education), dementia severity, diagnosis, quality of life, type of insurance
Caregiver: demographic info (race, age, gender, preferred language, education), self-efficacy
Both: caregiver relationship to patient (e.g., spouse, child), living status (i.e., whether patient lives alone, with caregiver, etc.), household income
Specifically, we plan to conduct the following exploratory analyses:
Explore whether SES (income, education) moderates the associations between pronoun use and caregiver outcomes
Probing the robustness of the associations between pronoun use and caregiver outcomes by including demographic covariates (i.e., caregiver age, gender, race, relationship to patient) and patient-related covariates (i.e., diagnosis, dementia severity, patient quality of life) as covariates in the primary linear regression analyses (described in Q5 above)
Test the correlation between I-word pronouns and caregiver self-efficacy
Test the correlation between pronoun usage and patient quality of life as well as patient dementia severity
Although the longitudinal survey data from this study have already been collected, we have not yet completed data reduction (i.e., coding pronoun use in the response to the “connection” question). We therefore consider this pre-registration to be valid, as we have not yet generated this key data.
