This Thread has easy example dataset! :D Help choose the right test please!

#1
We're trying to test whether there is a difference in length of visit (in minutes) between our original website and:
  1. Revamped Website #1
  2. Revamped Website #2

Number of different users = 13
The difficulty comes in because the data is unbalanced. Not all users visited each version of the website. The number of visits aren't always the same.

Here's the data! Copy + paste it to here: http://htmledit.squarefree.com
HTML:
<style type="text/css">
	table.tableizer-table {
		font-size: 12px;
		border: 1px solid #CCC; 
		font-family: Arial, Helvetica, sans-serif;
	} 
	.tableizer-table td {
		padding: 4px;
		margin: 3px;
		border: 1px solid #CCC;
	}
	.tableizer-table th {
		background-color: #104E8B; 
		color: #FFF;
		font-weight: bold;
	}
</style>
<table class="tableizer-table">
<thead><tr class="tableizer-firstrow"><th></th><th>Original Website</th><th>Revamp V1</th><th>Revamp V2</th></tr></thead><tbody>
 <tr><td>User 1</td><td>0.123568592</td><td>3.281797756</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>1.474941704</td><td>3.700737207</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>14.01709349</td><td>&nbsp;</td></tr>
 <tr><td>User 2</td><td>&nbsp;</td><td>0.453033244</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>35.43983492</td><td>&nbsp;</td></tr>
 <tr><td>User 3</td><td>&nbsp;</td><td>9.114522305</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>19.4191555</td><td>&nbsp;</td></tr>
 <tr><td>User 4</td><td>&nbsp;</td><td>2.857783866</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>0.964270145</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>37.47939464</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>26.79434689</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>24.93988443</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>0.997206803</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>42.27512141</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>96.55569889</td><td>&nbsp;</td></tr>
 <tr><td>User 5</td><td>3.019707122</td><td>37.32105432</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>9.968897982</td><td>88.84909681</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>17.53825342</td><td>74.29699071</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>47.49726499</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>26.52740119</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>27.91582575</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>71.54590594</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>90.6285037</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>68.27918192</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>19.12226641</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>88.47346066</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>77.34435825</td><td>&nbsp;</td></tr>
 <tr><td>User 6</td><td>&nbsp;</td><td>29.89688048</td><td>9.310827268</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>36.56776482</td><td>13.57866359</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>22.51034029</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>42.12484053</td><td>&nbsp;</td></tr>
 <tr><td>User 7</td><td>&nbsp;</td><td>25.73845621</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>25.09799092</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>22.43150293</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>56.12242273</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>73.52167695</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>58.48225946</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>19.21344262</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>64.45026218</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>0.400779611</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>7.572720877</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>13.82949972</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>91.97666207</td><td>&nbsp;</td></tr>
 <tr><td>User 8</td><td>&nbsp;</td><td>48.6898576</td><td>3.482786457</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>129.9262483</td><td>22.30556624</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>67.71332543</td><td>18.65564068</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>17.08463374</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>52.80661936</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>11.60592083</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>40.65997237</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>12.29300823</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>24.81482657</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>4.246564372</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>4.079373355</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>1.510015527</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>30.44575451</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>3.015585914</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>13.81682108</td></tr>
 <tr><td>User 9</td><td>1.892088286</td><td>31.59715495</td><td>22.92808124</td></tr>
 <tr><td>&nbsp;</td><td>8.477560538</td><td>49.72109156</td><td>38.19877336</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>30.2997121</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>14.52579414</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>40.10034129</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>14.48508491</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>0.586573698</td></tr>
 <tr><td>User 10</td><td>13.08722768</td><td>31.10623854</td><td>35.12534877</td></tr>
 <tr><td>&nbsp;</td><td>13.5526947</td><td>21.19729275</td><td>87.76317911</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>47.53907797</td><td>4.831172596</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>39.43917183</td><td>18.84919128</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>&nbsp;</td><td>8.79099872</td></tr>
 <tr><td>User 11</td><td>&nbsp;</td><td>24.98001138</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>26.29841929</td><td>&nbsp;</td></tr>
 <tr><td>User 12</td><td>&nbsp;</td><td>13.5559714</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>108.8702102</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>97.37861409</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>28.22666128</td><td>&nbsp;</td></tr>
 <tr><td>User 13</td><td>&nbsp;</td><td>48.18312037</td><td>29.81224447</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>69.74963166</td><td>34.3465346</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>34.46029381</td><td>23.57347001</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>36.35642741</td><td>25.33400863</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>110.548611</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>12.477009</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>178.5165748</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>137.7967765</td><td>&nbsp;</td></tr>
 <tr><td>&nbsp;</td><td>&nbsp;</td><td>168.0581469</td><td></td></tr>
</tbody></table>
Copy + paste the data above to here: http://htmledit.squarefree.com

How do I test if there's a difference in the length of visit (visit length measured in minutes) between the original website and revamp #1 or #2? The easiest way of course would just be to take averages across original vs revamp #1 vs revamp #2. But, considering these values come from different known users, there may be a better method. Help please! :D

Thank you! :D
 
#2
Cute html editing web site. I have made a note of it!

You have only 4 vistors to your original web site. That's unfortunate, because it is probably the one you want to compare to. I would probably do these in two ways, to show consistency. For both, I would get an average across all visits for a given subject to a given web site. Unless you want to get into fancy stuff like GEE, you can't let the individual visits count as independent. Only the subjects are independent. Then:

1. Oneway non-repeated ANOVA, ignoring paired data. You lose the within-subjects reduction in error, but increase the sample size.
2. Paired t-tests between the 3 web types, 2 at a time, but only using subjects that saw both.

Looks like the original website will come out low by either method.