Skip to main content

Table 1 Predictor variables and how they have been operationalised across the data sets

From: Demographic predictors of wellbeing in Carers of people with psychosis: secondary analysis of trial data

PredictorDescription
AgeAge here refers to the carer’s age at the time of the assessment. One study collected age as a categorical variable. Age was operationalised as either (a) Under the age of 50, or (b) Aged 50 or over (Dummy Codes: Under the age of 50 = 1; Aged 50 or over = 0).
GenderThis predictor refers to the carer’s gender. Gender was operationalised as either (a) Male, or (b) Female (Dummy Codes: Female = 1; Male = 0).
EthnicityMost of the carers identified as White British. All other ethnic groups were pooled together in a group we have labelled ‘ethnic minorities’. We therefore separated ethnicity as either (a) White British, or (b) Ethnic Minorities (Dummy Codes: Ethnic Minorities = 1; White British = 0).
Employment StatusWe divided carers as either (a) Currently Employed, or (b) Not in Employment. The first of these includes carers who were engaged in any amount of paid or voluntary employment; and the later includes carers who were unemployed, claiming unemployment benefits, on prolonged sick leave, retired, or not in work because they are a full-time carer (Dummy Codes: Not in Employment = 1; Currently Employed = 0).
Educational AttainmentCarers’ level of educational attainment was grouped by whether the carer had achieved a University-level qualification or not. That is, either (a) Higher Education, or (b) No Higher Education, respectively (Dummy Codes: No Higher Education = 1; Higher Education = 0).
Relationship to CRThere were a wide variety of carer-care recipient relational dynamics across the two trials. To reduce the number of groups we clustered carers by whether they were a (a) Parent to the CR, or (b) Other Relationship to the care recipient (Dummy Codes: Other Relationship to CR = 1; Parent to the CR = 0).
Care recipient DiagnosisThe term ‘psychosis’ in these trials was used in its broadest sense. We grouped carers based on whether they were caring for someone with a medically-defined psychosis diagnosis i.e. schizophrenia or schizoaffective disorder; or all other psychosis-related diagnoses. That is, either (a) Schizophrenia Spectrum, or (b) Other Psychosis Diagnosis (Dummy Codes: Other Psychosis Diagnosis = 1; Schizophrenia Spectrum = 0).
Living SituationWe separated carers according to whether they were living with their care recipient at the time of the baseline assessment, or not. We conceptualised this as either (a) Living with CR, or (b) Living separately from care recipient (Dummy Codes: Living with CR = 1; Living Without CR = 0).
Relationship StatusWe divided carers are either being (a) In a relationship, or being (b) Single. The first of these groups included any carers who were married, in a civil partnership, or cohabiting. Whereas, the second of these groups included carers who were single, separated or divorced (Dummy Codes: Single = 1; In a Relationship = 0).
Duration of CaringThis was the only continuous predictor included in our model. The duration that participants had been providing care was conceptualised as either the time since the care recipient psychosis onset, or an explicit report of how long theyhad been a carer for – both measured in months.