/sci/ - Science & Math » Thread #12566891

28KiB, 1147x257, Statistically Significant.png

View Same Google iqdb SauceNAO

SPSS - Statistically Significant?

Anonymous Mon 11 Jan 01:49:57 2021 No.12566891 View Reply Original Report

Quoted By: >>12567142 >>12567172 >>12567434

I have done an SPSS analysis to look at if different variables have a significant correlation in predicting depression.

The variables (pictured left) as you can see are Self Realisation, Pleasure, Autonomy and Control.

The significance value are placed under the Standardized Coefficients Beta. I am honestly having a really bad brain fart right now but are the scores of

-.104
-.187
-.164
-.160

Statistically significant or not?

Anonymous

Anonymous Mon 11 Jan 2021 02:52:12 No.12567014 Report

Quoted By:

Bump

Anonymous

Anonymous Mon 11 Jan 2021 03:34:24 No.12567135 Report

Quoted By:

Please respond

Anonymous

Anonymous Mon 11 Jan 2021 03:35:53 No.12567142 Report

Quoted By: >>12567152 >>12567160

>>12566891
I can't read your table. Are these quantitative or qualitative variables?
Looks quantitative so use ANOVA

Anonymous

Anonymous Mon 11 Jan 2021 03:37:15 No.12567152 Report

Quoted By:

>>12567142
It's quantitative.

Anonymous

Anonymous Mon 11 Jan 2021 03:39:14 No.12567160 Report

Quoted By: >>12567187

>>12567142
I am using the multiple regression as to see the difference in how the left variables predict depression. From my understanding the ANOVA is meant to be used only when looking at Statistical differences among the means of two or more groups.

Where I want to see which variables have a stronger prediction of depression.

Please don't leave me! I am only a sentence away from submitting my assignment, I just need to know if the

>STANDARDIZED COEFFICIENTS BETA
indicates statistical significance or not.

Anonymous

Anonymous Mon 11 Jan 2021 03:44:20 No.12567172 Report

Quoted By: >>12567195 >>12567203

>>12566891
Your p value tells you the significance of the result, which is the area under the curve relative to the positions of your t value. That's what you need to put, your p value not your t value.

It looks like all your correlations turned out significant, but if this is some social """science""" field then there's a 90% chance you've already fucked up your experimental design, making the result worthless.

Anonymous

Anonymous Mon 11 Jan 2021 03:47:57 No.12567187 Report

Quoted By: >>12567195 >>12567211

>>12567160
Oh ok. I can't really see your data. The image comes up too blurry on my computer. Multiple regression makes sense over ANOVA.
Yeah the standardized coefficient which SPSS calls beta tells you the strength of the effect of the independent on dependent variables.

Anonymous

Anonymous Mon 11 Jan 2021 03:51:27 No.12567195 Report

Quoted By: >>12567211

>>12567172
>>12567187
You prob know this but the CI can indicate if it's statistically significant depending if 0 is within the range or not for the standardized coefficient
SPSS should've have calculated the t-values for your regression analysis and given you a p-value. Sometimes it says it as sig.
Are you using 95% CI? Then if your "sig" is less than .05 you're good

Anonymous

Anonymous Mon 11 Jan 2021 03:53:34 No.12567203 Report

Quoted By: >>12567219

>>12567172
THAT'S ALL I NEEDED TO KNOW! THANK YOU!

>Social science
Yes it is psychology. I LOVE YOU ANON! THANK YOU! KILLED MY ANXIETY!

Anonymous

Anonymous Mon 11 Jan 2021 03:55:36 No.12567211 Report

Quoted By: >>12567219

>>12567187
>>12567195
Thanks so much anon. I was looking up the stupid internet to try to make sense of it but I couldn't Now that I know it is significant I can dump that assignment finally and submit it. THANK YOU SO MUCH!

Anonymous

Anonymous Mon 11 Jan 2021 03:57:41 No.12567219 Report

Quoted By: >>12567233

>>12567211
You're welcome.
>>12567203
>Yes it is psychology
Don't downplay yourself. Do some good research and replicate your findings. Learn some stats beyond your psych stats class. If pure statistics is too daunting, try applied or biostatistics. It will help you in your career.
Best of luck.

Anonymous

Anonymous Mon 11 Jan 2021 04:02:11 No.12567233 Report

Quoted By: >>12567241

>>12567219
Cheers man and happy belated new year. Stay safe during these times.

Anonymous

Anonymous Mon 11 Jan 2021 04:04:42 No.12567241 Report

Quoted By:

>>12567233
Thanks. You too.

Anonymous

Anonymous Mon 11 Jan 2021 05:51:49 No.12567434 Report

Quoted By:

>>12566891
The "Sig" column shows the p-value of your estimated coefficients. It looks like all of your variables have an effect on depression score that are significantly different from 0.

Anonymous

Anonymous Mon 11 Jan 2021 12:10:22 No.12568333 Report

Quoted By: >>12570116 >>12570133 >>12570149 >>12571411

Diff anon who would appreciate some statistics help.

I'm playing a game and trying to figure the chances of an event occurring each day, given multiple opportunities for the event to occur.

I have eight totems that have a 5% chance of producing my desired drop. My ignorant mind would think that the odds of it being produced would be 8 * 5% = 40% chance of dropping. But simply adding the chances together didn't seem to make sense to me.

Awhile back I was told to multiply the chances of each attempt failing instead. So, 0.95 ^ 8, which would yield a 66% chance of nothing appearing, and a 34% chance of yielding at least one instance of the drop that I was wanting. Am I stating right?

I was curious about what the odds would be if two totems produced instead. I would need to multiply 5% by 5%, resulting in 2.5%. The odds that two do not produce would be .975 ^ 8, approximately 81%. Which means that there is a 19% that two totems will produce in a specific day. My initial impulse was to add these two together, 34% + 19%, but that does not make sense as I think about it because it would exceed the odds that at least one would produce.

Am I right here, or just totally off-base?

Anonymous

Anonymous Mon 11 Jan 2021 21:31:20 No.12570116 Report

Quoted By:

Anonymous

Anonymous Mon 11 Jan 2021 21:36:03 No.12570133 Report

Quoted By:

>>12568333
You should really post this sort of thing in the stupid questions thread so people actually see it. Anyway, you're repeating multiple independent experiments (dropping totems without totems affecting the other) whose results are binary (getting desired drop or not). Look up the binomial distribution. The joint probability of multiple independent events is the product of the probability of the events, so the probability of all 8 experiments being failures is just 0.95^8. The probability of at least 1 experiment being a success is just 1 - 0.95^8 (1 minus the probability of all failures). If you redefine success as two totems getting the desired drop, then the probability of success is indeed 0.025. The probability of all drops being failures is 0.975^8 (assuming you're still doing 8 trials) and the probability of at least 1 success is then 1 - 0.975^8. Use the binomial distribution probability mass function to get the probability of at least two successes (1 minus probability of no success minus probability of one success will give you the probability of at least two successes).

Anonymous

Anonymous Mon 11 Jan 2021 21:41:00 No.12570149 Report

Quoted By:

Anonymous

Anonymous Tue 12 Jan 2021 05:07:55 No.12571411 Report

Quoted By: >>12571412

>>12568333
>So, 0.95 ^ 8, which would yield a 66% chance of nothing appearing, and a 34% chance of yielding at least one instance of the drop that I was wanting. Am I stating right?
This is correct.
>I was curious about what the odds would be if two totems produced instead.
The fixed odds of having the first two totems drop and the rest not are calculated by (0.95^6)(0.05^2) = 0.0018.
Now since the order doesn't matter, there are 8 choose 2 possible orders of your 2 drops and 6 non-drops.
The probability of exactly 2 drops is:
(2C8)*(0.95^6)*(0.05^2) = 0.0515, about 5%

The above formula is just the probability mass function of the binomial distribution.

Anonymous

Anonymous Tue 12 Jan 2021 05:08:57 No.12571412 Report

Quoted By:

>>12571411
That's (8C2) not vice versa, small typo

Capcode	All Only User Posts Only Moderator Posts Only Admin Posts Only Developer Posts
Show Posts	All Only With Images Only Without Images
Deleted Posts	All Only Deleted Posts Only Non-Deleted Posts
Ghost Posts	All Only Ghost Posts Only Non-Ghost Posts
Post Type	All Only Sticky Threads Only Opening Posts Only Reply Posts
Results	All Grouped By Threads
Order	Latest Posts First Oldest Posts First

Your latest searches