Correlation analysis Univariate

#1
Hello fellow,

I recently joined the statistical forum, it is time to join as statistical-member to this forum since my carrier progress will be around researching or/and seeking similar interests.

through a research about cybersecurity, I have a sample of 60 participants of companies where data survey collected depicts CEO gender, CEO security experience, $Value investment on security, No of security breach attempt, No. of successful security breach attempt, age of company, security rating.

The hypothesis informs that more investment on security should tend to have less successful breach attempt, whereas less investment tend to have more successful attempt.

Ha: Companies investing more on security tend to have less successful attempt than those companies who don't.

test is to identify the correlation between ID: $Value_Invest and DV: No_Success_Attempts

SPSS univariate analysis shown a correlation between the two variables $Value investment and Successful attempts with r > 0 and p-value < .001. great, variables are correlated but positively. this means if we invest more we will tend to have more successful attempt? while my expectation is to get a negative correlation instead.

Any help?
 
Last edited:

Karabiner

TS Contributor
#2
In our garden, we grow wonderfully smelling herbs. Unfortunately, these tends to attract snails. Therefore, after haveing suffered from snail attacks, we invested money in poison, fences etc. This helped much, but we have still a lot snails in our garden.

Our neighbour doesn't have snail-attracting herbs in his garden. He has nearly no snails. He has not invested into anti-snail-defenses.

Would one therefore conclude that poison and snail fences attract snails?

There possibly are at least 2 severe problems with your design, analysis and interpretation:

a) a cross-sectional survey with observational data is used for causal interpretation. You should have 2 data points at least, in a longitudinal design.

b) you left out important explanatory variables, e.g. the companies' respective attractiveness for attackers.

With kind regards

Karabiner