How to create feedback

silviabradatsch edited this page Sep 2, 2015 · 14 revisions

How to…

…generate and plot Feedback (under construction)

If you want to generate Feedback that gives your participants an impression of how they scored, you need the program (R) to identify and aggregate the intended scales of items. There are certain conventions that make this step as easy as possible.

Naming the items

When naming your items pay attention to the following conventions:
  1. Make sure it´s clear which item belongs to which test and scale. For example, if your first item is an item of the extraversion-scale of the Big Five Inventory (BFI), you should name this item BFI_extra1. The second item would be BFI_extra2 and so on.
  2. If your item is reversed and therefore the results of this item have to be converted, you should mark this item by adding R to the item-name. Consequently, if the first extraversion-item was reversed, it should be named BFI_extra1R.

If you consistently follow these conventions, the program will be able to identify your scales on its own.

Plotting Feedback

The next step is plotting the Feedback in formR so it can be viewed by the participants of your study. The following text is an example that shows how to tell formR to generate and plot Feedback.

opts_chunk$set(fig.width=10,fig.height=8)
library(ggplot2); library(reshape2); library(car); library(formr)

## automatisch Daten & Zahlen und so richtig erkennen
Bindungsbefragung = formr_recognise(results= Bindungsbefragung , item_list = NULL)
# automatisch Skalen bilden, items mit R am Ende werden automatisch umgedreht
Bindungsbefragung = formr_aggregate(results=Bindungsbefragung, item_list = NULL)



Ihre Bindung zum Partner

Bindungsbefragung$Sicherheit = (Bindungsbefragung$BBE_sicher - 4.35) / 0.53
Bindungsbefragung$Abhaengigkeit = (Bindungsbefragung$BBE_abhaengig - 2.82) / 0.68

Bindungstyp = melt(Bindungsbefragung[,c("session", "Sicherheit", "Abhaengigkeit") ], id = 'session')

qplot_on_bar(Bindungstyp, "Ihr Wert", "Eigenschaft") 



qplot_on_normal(Bindungsbefragung$Sicherheit, xlab = "Sicherheit", ylab = "Prozentzahl anderer Menschen mit diesem Wert")

With these instructions, formR plots a normal distribution and marks where your participant is placed on the distribution. This will look like the picture shows:

The blue line shows the participant that he/she is more extraverted than most of the population.

Written Feedback

The following expression creates Feedback in a standardized pattern, depending on the range of standard deviation your participants score is in.



Diese Grafik zeigt Ihnen:

r feedback_chunk(Bindungsbefragung$Sicherheit, c("Sie sind ängstlich gebunden.","Sie sind eher ängstlich gebunden.", "Sie sind durchschnittlich sicher gebunden.", "Sie sind eher sicher gebunden", "Sie sind sicher gebunden."))