I am in a statistical math class and I have a project to do. So I would like to ask for a small hand from my friends here. I am doing a study and developing a hypothesis based on computer reboot times.
All computers used for my study must be newer(ish), no old sub 1000mhz systems if you would please. Newer single core or dual core (2000+ish MHZ) would be most ideal; Windows only for this hypothesis.
Simple procedure, get out yur stopwatch (or watch with a second hand) close all programs running (not task bar items) hit the start menu and press restart then start your timer.
Please post your time to a complete reboot (total minutes and seconds), to the point where you can actually start a program, not just the splash screen. If you wouldn't mind tell me your CPU and RAM amount (hard drive size/type is not important for this base line).
You should be able to tell when your computer is finished booting, HD activity slows way down, and program/The Start Menu should open easily. Try to be as exact as possible, do it twice to make sure you have an acurate time.
This is a team project but my team is slacking off and I need more than 30 numbers for a large sample size. If you see 30 responses to this then I have enough numbers for my hypothesis.
Ok, it takes about 35s just for the shutdown, 1m10s more to get to the login screen, 15s from login to desktop and 10s more to have Firefox running. I did it twice and got similar results more or less 1-2 seconds. So it takes me 2m-2m10 for a full reboot.
We have a small range of older systems. All with varying degrees of crap installed on them. With reboot times that range from ~6 mins to ~15mins.
I wanted a range of newer systems with a completely random sample. So there would be a wide range of crap installed but would be a fair sample of the average (Newer) computer system.
I will be using hypothesis testing on the mean for both a small sample size (older systems) and a large sample size (newer systems). Standard deviation will be determined (of course) and a conclusion will be drawn based on the Null/Alternative Hypothesis for each case. All done in the name of showing that newer systems would be more productive to a certain degree, than the older systems. This is not a hard core (BE ALL END ALL) direct comparison between a single old system and a faster new system. So this is a perfect type of test for what my assignment requires. Remember this is a Statistics Class, not a Computer Class.
HDD size/type does not matter because it will only give a very fast number (or two) out of a random sample, which would actually be ideal for this test, because a full spectrum of what is available (for newer computer) will show up in the sample.
Wish I would have seen this thread earlier, as I woulda thrown in some results for you as well. Never the less, if you remember to and have enough time, post up the results of your study- I'm curious to see the results. Don't worry about translating the stats., had to do plenty of that crap in school myself.
These aren't real benchmark results and by design the numbers don't say jack-crap lol
Actually the interpretation is the boring part. Yep old systems are slow as hell, new systems seem like speed demons in comparison.
The mean of the newer system restart times is ~354% faster, that's the interesting number. I wonder if real world, productivity benchmarks show a common newer computer as 3.5 times faster than an older (common for the time period) 900mhz with 512mb of ram...
If you are wondering about all the stats..
Older Systems = CPU = 900 MHZ, 512MB of RAM
Mean restart times = 496.17 seconds
Standard Deviation = 141.16 seconds
Standardized Test Statistic: t= 3.4 @ 0.01 level of significance
I used a t-test of the mean for small sample sizes
Newer Computer Systems
Mean restart times = 140.37 seconds
Standard Deviation = 65.82 seconds
Standardized Test Statistic: z= 14.75 @ 0.01 level of significance
Used a z-test of the mean for large sample sizes
Had to build a hypothesis on that - and it was a very simple test at that....