1
$\begingroup$

I am new to statistics and I need some help in understanding how to report the data of some tests I am running on R, I hope this is the right place!

I have a dataset:

# A tibble: 6 x 12
  `Participant Pr… Group Gender   Age `Spreadsheet Ro… Screen_Name Condition Reaction_Time Correct_Humorous
  <chr>            <chr> <chr>  <dbl>            <dbl> <chr>       <chr>             <dbl>            <dbl>
1 CG_01            cont… F         18                2 Joke        Differen…         3156.                1
2 CG_01            cont… F         18                3 Joke        Differen…        15219.                1
3 CG_01            cont… F         18                4 Joke        Differen…         4005.                1
4 CG_01            cont… F         18                5 Joke        Differen…        30027                 0
5 CG_01            cont… F         18                6 Joke        Differen…         9811.                1
6 CG_01            cont… F         18                7 Joke        Differen…        12106.                1
# … with 3 more variables: Wrong_Related <dbl>, Wrong_Unrelated <dbl>, Answer_Type <chr>

I am trying to look at accuracy between two groups of subjects in two conditions.

So I ran a linear model

model_acc_ch <- lm(data= data4, Correct_Humorous ~ Group + Condition)

I got this result:


Call:
lm(formula = Correct_Humorous ~ Group + Condition, data = data4)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.95739  0.04261  0.13864  0.32987  0.42590 

Coefficients:
                         Estimate Std. Error t value Pr(>|t|)    
(Intercept)               0.86136    0.01556  55.357  < 2e-16 ***
Groupdyslexia            -0.28726    0.01840 -15.613  < 2e-16 ***
ConditionSame word class  0.09603    0.01836   5.231 1.88e-07 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.3893 on 1796 degrees of freedom
Multiple R-squared:  0.1311,    Adjusted R-squared:  0.1302 
F-statistic: 135.5 on 2 and 1796 DF,  p-value: < 2.2e-16

I want to write down the result, it's my first time and I don't know how to report the F values, so I ran an Anova on the model that looks like this:

Analysis of Variance Table

Response: Correct_Humorous
            Df  Sum Sq Mean Sq F value    Pr(>F)    
Group        1  36.931  36.931 243.676 < 2.2e-16 ***
Condition    1   4.147   4.147  27.365 1.881e-07 ***
Residuals 1796 272.196   0.152                      
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

I don't know what might be more significant. would it be ok if I say "Analyses of variance (ANOVAs) showed significant differences for the number of correct choices between groups [F= 243.676, p < 0.0001] and between conditions [F=27.365, p<0.0001]"?

Thank you to anyone who can help!

$\endgroup$
  • 1
    $\begingroup$ You need a more precise title, something more like "Reporting ANOVA results". I would always want to see means mentioned in a report from ANOVA, because that is what the procedure compares. $\endgroup$ – Nick Cox Aug 21 at 6:47
  • $\begingroup$ You did not run what is commonly called a linear mixed (effects) model $\endgroup$ – Roland Aug 21 at 10:33
0
$\begingroup$

As I understood, you have two binary independent variables named Group and Condition and one binary dependent variable named Correct_Humorous. You are going to need a glm with the binomial family and logit link:

fit <- glm(
  Correct_Humorous ~ Group + Condition,
  family = binomial(link = logit),  # logistic regression
  data = data4
)

summary(fit)

fit$rsquare <- 1 - (fit$deviance/fit$null.deviance)

Moreover, to find out the effect size of the independent variables on the outcome, you can calculate odds ratios (OR) with confidence intervals (CI):

install.packages("questionr")
require("questionr")
odds.ratio(fit)

In such scenarios, you need to report goodness of fit (such as AIC, $R^{2}$), the Estimate or OR with 95% CI and p-values of the corresponding fit.

If you have more questions or problems, you can provide a reproducible example and ask on StackOverflow.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.