Keyboard Shortcuts
Likes
Search
Re: confused with the answer from the calculation of UCINET [1 Attachment]
开云体育Jen, try ? ->dsp centdiff(padgm padgb undir) ? As for densitydiff, as I explained before, your version of the program is calculating things correctly but is printing the wrong label: in the output, it is using the word centralization instead of density. This has been corrected, but you would have to download the current version to see it. ? steve ? From: ucinet@... <ucinet@...>
Sent: Thursday, March 21, 2019 08:12 To: 'Steve Borgatti' steve.borgatti@... [ucinet] <ucinet@...> Subject: Re: [UCINET] confused with the answer from the calculation of UCINET [1 Attachment] ? ? Dear Professor Borgatti: Thank you for your information about my question! I try to follow the steps you suggest. However, it always come up with the problems.? Can you help me to solve the problem again? For example, I try to compare the centralization of two different networks (logis 1 and logis 2). The analysis result come up with as follows which does not show any number to tell the analysis of difference. Is there any process wrong? In addition, when I try to compare the density of two different networks, the software keeps showing me the wrong process message. I just don't understand what the problem is. Can you help me again!?? Thanks a lot! ?
? 任庆宗 感恩合十 ======================================= Department of Logistics Management,? National Defense University ? ? 在 2019年3月20日 星期叁 上午10:15:16 [GMT+8]， 'Steve Borgatti' steve.borgatti@... [ucinet]<ucinet@...> 写道： ? ? ? Jen, first, you’re right: it’s mislabeled. But it is doing density. As for the numbers, is it possible your data are valued? If so, the program will use average tie strength. If your hand-calculating was based on dichotomized data, you would get different answers. You could do ? ->d1 = dichot(network1rows) ->d2 = dichot(network2rows) ->dsp densitydiff(d1 d2 undir) ? steve ? From:
ucinet@... <ucinet@...>
? ? Dear All: I try to compare with the densities of two different networks. The result come as followings: ? ->dsp densitydiff(Network1Rows Netork2Rows undir) ? Bootstrap SE1 = 0.1354 Bootstrap SE2 = 0.2363 T-statistic? ?= 1.3690 Cohen's d? ? ?= 1.9163 ? Test for equal degree centralization. 20000 iterations. Data are undirected. ? ? ? ? ? ? ? ? ? ? ?1? ? ? ? ? ? ? ? ?Measure? ? ? ? ? ? ? ? ?-----? ? ? 1? ?Network1 ROWS 0.506? ? ? 2? ?Network2 ROWS 0.875? ? ? 3 Obs Diff 0.369? ? ? 4? p-value 0..090? 4 rows, 1 columns, 1 levels. ? ? I have two questions about the result. First, the title of the result "Test for equal degree centralization.20000 iterations. Data are undirected." makes me a little confuse because I am testing the difference of density not the centralization. Why the title shows the test for equal degree centralization? Secondly, it looks like that there is no significant difference between the two networks in terms of density. My problem is the number of density calculated by UCINET is different from the number by hand calculation. According to the result, the density of the? Network2 is 0.875. However, my hand calculation is only 0.0708. I don't understand what the problem come from. Can any one help me solve this confused problem.? Thanks a lot! ? ? ? ? ? ? ? ? ? ? ? Jen
|