# What Statistical Tests should I run

#### PsychDummy

##### New Member
What statistical test should I run for the following study

Study 1:
STUDY 3:

Reasearch Question: Adults compared to children might adopt more sophisticated eye-following behaviours (i.e., better at looking the eye, and for longer times) as they are more experienced and advanced in communications and other forms of social interactions.

Aim: To explore whether the following gaze behaviours might differ from people from different age groups.

Main effect of image type: A larger proportion of fixations on the face for the Human than the
monster images.
Main effect of age: children might spend a larger proportion of fixations on the faces than adults
because of curiosity,
Interaction between image types: difference in the mean proportion of fixations on the face would
be larger for university students than that of the primary school children due to more sophisticated
eye-following behaviours.

Design:
Between-subjects design
2 (image types: Human vs. Monster) × 2 (age group: Primary School vs. University
Students).
Demographic:
Participants were recruited from two age groups: primary school students and university
students
80 participants total
20 Participants within each image and age group

2 IVs:
Human
Monster

6 conditions: (run on STATA)
Gender = Participants' gender
Age = Participants' age
Image = type of image
Ssgroup = the age group
Condition = unique combination of the image type and the age group
Faceperc = the mean proportion of fixations on the face across 36 images

#### PsychDummy

##### New Member
Also, I'm not entirely sure what the independent and dependant variables are, can someone confirm.

#### priyac1987

##### New Member
If your aim is to check how age affects gaze behaviour, then AGE is the independent variable and GAZE BEHAVIOUR is the dependent variable. Regression analysis is the best fit for your case.
The other 5 conditions (gender, image, ssgroup, condition and faceperc) seem to be control variables. You can apply multiple linear regression test with these as the control variables. SPSS is friendly for your case.