CRIME FIGHTING IN A NUTSHELL

 

Hi Folks!! I am back again with Chi-Square test to break the multitudes opinion about world being unsafe everyday. Also, I have done Coefficient of Variation test to break this myth. 

Step 1: I did a Chi-Square test to find out on what dates crimes occur the most. I found out that many police officers believed that, days close full moon (before/after), crime increases. So, I created a schedule in my model based on dates of the full moon and I naturally started off with full-moon day as my first day like below (1st of every month).

As you can see from above, 1st of month is presumably the full moon day and number of crimes occurred are highlighted in orange. Like that I repeated the same for 3 months like below

In the next column, since people assumed that danger is omnipresent everyday, I divided the 90-days crimes by 90 which gave 6 as a residual. Then, I carried forward calculating f0-fe, (f0-fe)^2 and (f0-fe)^2/fe to finished off my required calculations for this test. 

The results are just as I imagined, criminal intent rises amongst individuals only on particular days, but not on everyday. This Hypothesis is formulated here to prove that everyone's belief is wronged by me. 

'H0:B0= Crimes are committed equally on everyday' 

This leaves the Alternative hypothesis: Crimes can occur sporadically or concentrated, for which I have calculated Standard Deviation to find out the models dispersion of crimes from their total mean like below.

I took great care in calculating Class Interval Mid Values in Row I. After which, all the functions worked properly because the mean relies upon this and the models accuracy depends upon this. The results of this test are like below

The test results are simple to interpret, SD is higher than 1 showing that the correlation is weak to minimise dispersions of crime occurrences points and our objective. I applied COV on it (SD/MEAN) and it gave a 8917% variability in results representing what happened was due to various factors and cant pinpoint the moon phases exactly as guilty.

 

But, I am inclined to emulate through my own means what other empiricist psychologists who manifested that waxing and waning are the culprits. Hence, I applied SD & COV formulas for individual frequencies (f) and here are the results

The aforementioned results also hold evidence that highest crimes (65% of total) were recorded during the frequency 1 to 3 & 28 to 31. For further clarity, the above data proves that COV is nil at 1% only for this frequency we are talking about right now.  This 1% is in contrast to 8917% variability and has more consistency in reporting that crimes are concentrated between these dates.

Sorry, I had no time yesterday to test COV further. Today, on 19-11-2021, I have taken my previous vaccine data and experimented on attaining best results by choosing MEAN from grouped and ungrouped data. The results were like below and ungrouped SD formula has better accuracy than sample means of grouped data for this model. Here, the COV has no variation. I found out, this method could prove useful to identify single observations strengths with other observations. Here, the objective COV is ZERO. 

I hope you all concur to my viewpoints here and even if you like to criticise this model, just rememeber, this is easy for anyone to monitor like a log. Just input past crimes data->Run the Hypothesis-> If yes, then find out COV %. If it is safe, relax but keep monitoring further.

Beware of the ware wolves and Buffy the Vampire on a full moon day LOL!!

Download a copy here

This is patented and copyrights belong to me.