Help with analysis. PLEASE.

Help with analysis for my psych dissertation. PLEASE.

Hi all:wave: , please can any one give an indication of which stats test to use. So far I've looked at doing a bivariate and found significant data, but I need to find a pattern between alcohol consumption and personality/stress. Perhaps a MANOVA?

Independent Variable; Alcohol consumption (7 factors, each day of the week)
1st Dependant V; Stress Score (1 factor)
2nd Dependant V; Personality Type ( 5 factors, personality type i.e. neurotic, extrovert, introvert etc.)

Any input would be hugely appreciated. I'm using SPSS.
Last edited:


New Member

It seems that personality and stress should be the independent variables, and alcohol should be the dependent variable. Personality and stress are likely to influence one's level of alcohol consumption rather than alcohol influences personality and stress (we have a personality from young).

How did you measure alcohol as seven factors? By amount of alcohol drank?
Thanks for the reply. It was measured per unit of alcohol, with each day of the week being a factor. e.g.

mon- 5 units
tue- 11 units
wed- 3 units........

thanks george


New Member
What you should do is add the amount of alcohol for the 7 days, and divide by 7 to create mean amount of alcohol consumption (i.e. one variable) for each person, and use this as a dependent variable. Personality and stress can act as predictors and alcohol consumption as the dependent variable.

How was personality measured? in terms of scores for separate categories such as extraversion,neurotic, etc, or was just categorical..where one person was just given a personality type.

Tell me if you recognise the independent and dependent variable classification for alcohol consumption (as Dependent) and personality and stress (as Independent)
thanks again for the reply.

Personality was measured on a likert scale (1 to 5) from 60 questions. The results from my 36 participants were put into a syntax file (designed by the lecturer) from SPSS, which resulted in 5 scores/personality facets for each individuals personality type (I.e. extravert 4.02, introvert, 1.1, openness 2.22). from 0.00 to 5.00

The title for my research is ‘Does alcohol consumption correlate to stress and personality type’. So I suppose alcohol could be considered a DV.

I could use alcohol as one variable, but I was told to find a pattern within the alcohol consumption - is this possible from a mean alcohol consumption per person. I hope this is clear

Thank you, George.
Last edited:


New Member
Hi G,

Alcohol consumption is definitely the dependent variable..I wont lead you wrong as this is my area as dont worry.

You can calculate the mean alcohol consumption per person per day by adding scores via the SPSS menubar in Transform, Compute: you add the scores separated by plus signs then divide by 7. This computation in SPSS will automatically compute the mean alcohol consumption score for each person (per day as you divide by 7). However you have only 32 participants, and multiple regression requires a large sample size. You can do correlations with all variables to see how each personality type and stress correlate with alcohol consumption.

In terms of pattern, I am not sure what you mean in this case? Explain
Thanks, thats useful stuff. So I've 'computed the variable'. I then used a bivariate to find significant data - which worked. Also an ANOVA. Which sort of statistical test would you use?? I'm I going along the right lines?

As you can gather i'm certainly no SPSS genius
Last edited:


New Member
You can use correlations to examine the relationship among all variables and specifically determine which variables alcohol consumption per day correlates with. You only do ANOVA if you have categorical (nominal) independent variables which I recognise you dont have, so ANOVA is not suitable here. Correlations are your best tools.