Regression-Based Normative Data for Children From Latin America: Phonological Verbal Fluency Letters M, R, and P

Abstract

This study is part of a larger project to generate norms for letter verbal fluency test (VFT) in 3,284 children from nine Latin American countries. The letter VFT (letters M, R, and P) was administered and multiple linear regressions, including age, age2, MPE (mean parental education), MPE2, sex, and interactions were used as predictors. Results showed significant differences across countries for all scores. Age affected scores linearly except for Ecuador (P-letter), in which a quadratic effect was found. Scores increased linearly as a function of MPE, with the exception of Mexico (R-letter), in which a quadratic effect was found. Age by MPE (M- and P-letters), and age by MPE2 (R-letter) interactions were found in Mexico. Sex had an impact on letter R in Cuba, and letter P in Ecuador and Paraguay. Age2 by sex interaction was found in Ecuador (P-letter). These norms will be useful for clinical neuropsychologists in these countries to evaluate their patients’ verbal fluency.

Normative Data for Test of Verbal Fluency and Naming on Ecuadorian Adult Population

“…The normative data that fit for the demographic variables were established through a four-step procedure, using the final regression model obtained at the end of the procedure (Rivera et al, 2019(Rivera et al, , 2020: (a) The expected test score (Ŷ i ) is computed based on the fixed effect parameter estimates of the established final regression model:…”

Phonological and semantic verbal fluency test: Scoring criteria and normative data for clustering and switching strategies for Colombian children and adolescents

“…Using beta values of each final regression model, four steps were followed (Rivera et al., 2021) to estimate z ‐scores $$( {{z_i}} )$$ and percentile $$( {P{c_i}} )$$ adjusted to demographical variables: (1) the predict score $$( {{{\hat Y}_i}} )$$ was computed based on the fixed effect parameter estimated of the established final regression model; (2) to estimate residual value $$( {{e_i}} )$$ the following formula: $${e_i} = \;{Y_i} - \;{\hat Y_i}$$ was used; (3) residuals should be standardized using the residual standard deviation (SD e ) of the regression model: $${z_i} = {e_i}\;/{\rm{S}}{{\rm{D}}_e}$$; and (4) to convert $${z_i}$$ to $$P{c_i}$$, the standard normal cumulative distribution function was used if the model assumption of normality of the standardized residuals was met in the normative sample. If the standardized residuals were not normally distributed in the normative sample, the empirical cumulative distribution function of the standardized residuals was used.…”

“…Multiple regression analyses were conducted for each clustering and switching strategy following Rivera et al (2021) methodology. The full regression models were included as predictors: age, quadratic age, sex, MPE, quadratic MPE, type of school and two‐way interactions between the fixed effects.…”

“…It seems that children and adolescents whose parents had higher educational level perform better in VFT compared with those with lower educational level (i.e., Olabarrieta‐Landa et al., 2017b; Rivera et al., 2021; Van der Els et al., 2011). Sex has been inconsistently associated with VFT; however, most studies have reported no sex differences (i.e., Malloy‐Diniz et al., 2007; Ruffieux et al., 2009; Tallberg et al., 2011; Van der Elst et al., 2011) or when differences arise, its impact was very small compared with age or parents’ education (Olabarrieta‐Landa et al., 2017b; Rivera et al., 2021). Thus, both age and parents’ education level are extremely important variables to consider when assessing verbal fluency in this population.…”

“…Due to its broad use in clinical and research settings, the number of studies focused on providing normative data for VFT in the paediatric population has increased during recent years (i.e., García et al., 2012; Martins‐Oliveira et al., 2016; Olabarrieta‐Landa et al., 2017b; Rivera et al., 2021; Van der Elst et al., 2011). However, most of these studies have focused on TS, and more normative data for clustering and switching strategies during VFT are needed, especially for Latin American Spanish‐speaking countries.…”