Skip to main content

Table 2 Adjusted R squared for multiple linear models using all, uncorrelated, and four available terms

From: Google Health Trends performance reflecting dengue incidence for the Brazilian states

State Adjusted R squared – All terms model (n) Adjusted R squared – Uncorrelated terms model (n) Adjusted R squared – Four terms model (n = 4)
Acré (AC) 0.126 (7) 0.110 (6) 0.092
Alagoas (AL) 0.481 (8) 0.418 (3) 0.470
Amapá (AP) 0.096 (7) 0.099 (6) 0.103
Amazonas (AM) 0.847 (9) 0.590 (2) 0.821
Bahia (BA) 0.647 (11) 0.603 (3) 0.624
Ceará (CE) 0.839 (10) 0.746 (3) 0.812
Distrito Federal (DF) 0.829 (10) 0.433 (2) 0.830
Espírito Santo (ES) 0.725 (10) 0.537 (3) 0.688
Goiás (GO) 0.785 (11) 0.660 (3) 0.768
Maranhão (MA) 0.859 (9) 0.615 (2) 0.856
Mato Grosso (MT) 0.573 (9) 0.453 (3) 0.559
Mato Grosso do Sul (MS) 0.713 (9) 0.583 (3) 0.694
Minas Gerais (MG) 0.923 (12) 0.718 (4) 0.919
Pará (PA) 0.600 (9) 0.277 (2) 0.596
Paraiba (PB) 0.837 (10) 0.683 (3) 0.832
Paraná (PR) 0.845 (12) 0.741 (4) 0.821
Pernambuco (PE) 0.819 (10) 0.714 (2) 0.808
Piauí (PI) 0.553 (8) 0.377 (3) 0.542
Rio de Janeiro (RJ) 0.765 (12) 0.704 (5) 0.744
Rio Grande do Norte (RN) 0.890 (9) 0.706 (2) 0.891
Rio Grande do Sul (RS) 0.804 (12) 0.724 (4) 0.779
Rondônia (RO) 0.568 (9) 0.415 (4) 0.544
Roraima (RR) 0.093 (7) 0.091 (6) 0.058
Santa Catarina (SC) 0.826 (9) 0.788 (2) 0.825
São Paulo (SP) 0.930 (13) 0.861 (6) 0.919
Sergipe (SE) 0.308 (8) 0.216 (3) 0.300
Tocantins (TO) 0.415 (8) 0.342 (4) 0.401
Brazil 0.888 (14) 0.852 (7) 0.850
  1. Different numbers of terms (n) were available depending on the state (e.g., São Paulo vs. Acré) and were correlated considering a state-by-state basis (e.g., Amapá vs. Maranhnão; Additional file 3: terms per state). Four terms model was developed with “dengue”, “dengue sintomas”, “aedes”, and “mosquito”. Bold represent adjusted R squared values above 0.8. Data for Brazil is shown in the last row