Repeated measures ANOVA

df <- read_csv("https://docs.google.com/spreadsheets/d/e/2PACX-1vTGUepm0ddAefpaILx5QR9eCBgvnY-st6FBfVf39W1meHwGGsXvb63iVhNFZ3DvwXCM9ZsFaVo4sdyp/pub?gid=1906853507&single=true&output=csv")

── Column specification ────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
cols(
  Tipo = col_character(),
  Tiempo = col_character(),
  `Viabilidad bacteriana` = col_number()
)

añadir una columna ID

Check

Outliers?

df %>% 
  ggplot(aes(x = `Viabilidad bacteriana`)) + 
  geom_histogram()

`stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

Cuál es outlier?

boxplot.stats(df$`Viabilidad bacteriana`)$out

[1] 21332

df %>% 
  ggplot(aes(x = `Viabilidad bacteriana`)) + 
  geom_histogram() + 
  facet_grid(Tipo ~ Tiempo)

`stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

summary(df$`Viabilidad bacteriana`)

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
-7268.0   676.2  4802.5  4976.1  8922.2 21332.0 

chequear el valor negativo

Identificar outliers por cada grupo

Chequear esos valores

hay NAs?

any(is.na(df$`Viabilidad bacteriana`))

[1] FALSE

Verificar esfericidad

PENDIENTE

Análisis

df %>%
  group_by(Tipo, Tiempo) %>%
  summarise(n = n(), 
            mean = mean(`Viabilidad bacteriana`), 
            sd = sd(`Viabilidad bacteriana`))

`summarise()` has grouped output by 'Tipo'. You can override using the `.groups` argument.

Model

summary(anovaModelRM)

Error: ID
     Df    Sum Sq   Mean Sq
Tipo  1 175634105 175634105

Error: Within
             Df    Sum Sq   Mean Sq F value   Pr(>F)    
Tipo          1 2.684e+05    268378   0.015    0.904    
Tiempo        6 8.077e+08 134616667   7.376 1.18e-06 ***
Tipo:Tiempo   6 4.166e+07   6943169   0.380    0.890    
Residuals   111 2.026e+09  18251093                     
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

The repeated-measures ANOVA (formula: Viabilidad bacteriana ~ Tipo * Tiempo + Error(ID)) suggests that:

• The main effect of Tipo is NA and very small (F(1, 111) = 0.01, p = 0.904; Eta2 (partial) = 1.32e-04, 90% CI [0.00, 2.97e-04])
• The main effect of Tipo is not significant and large (F(6, 111) = 7.38, p < .001; Eta2 (partial) = 0.29, 90% CI [0.15, 0.37])
• The main effect of Tiempo is significant and small (F(6, 111) = 0.38, p = 0.890; Eta2 (partial) = 0.02, 90% CI [0.00, 0.02])
• The interaction between Tipo and Tiempo is not significant and very small (F(1, 111) = 0.01, p = 0.904; Eta2 (partial) = 1.32e-04, 90% CI [0.00, 2.97e-04])

Effect sizes were labelled following Field’s (2013) recommendations.

Plot

Pendiente