|
Re: Number of Ties in NETDRAW
开云体育Hi Scott, I get different numbers than you, but in any case, in ucinet choose Transform|Dichotomize interactive and then press “load” button. Enter the 2-mode davis dataset (which I’ve called “women”). You should see this: ? ? The “number of 1s” column is probably what you want. ? Steve. ? ? ? ?
? From: [email protected] <[email protected]> On Behalf Of Scott Thomas via groups.io
Sent: Saturday, December 7, 2024 17:04 To: [email protected] Subject: [ucinet] Number of Ties in NETDRAW ? If one takes the Southern Women 2-mode data (14x18) and converts it to 1-mode and runs Netdraw, one gets 278 ties in the network. As one removes ties, the pattern is: 278 ,194, 104,, 56, 24, 20, 6, 0? Is there a way besides Netdraw to obtain a printout of these same values?? Thank you. |
||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Help Reg. Partitioning Data
Hello All,
?
I am performing a Weighted Correlation Network Analysis (WCNA) where the correlation coefficients across a set of variables are used as the tie strength for their correlational network. I have performed a Principal Component Analysis on the data (in SPSS) to get the factor structure of the variable set. The PCA result shows that the 31 variables can be grouped into 8 components.?
?
Now I want to run UCINet to perform the following:
I understand that a partition file needs to be created with the 31 variables and their respective attributes. Somehow, I am unable to execute the process.?
?
Please help me with the steps!
?
Thanks in advance,
?
Aritra. |
||||||||||||||||||||||||||||||||||||||||||||||||||||||
开云体育Clark, you’re welcome to send me your data (sborgatti@...). But first you might check that all of the matrices have normal values. Sometimes people will have 1s and missings instead of 1s and 0s, which won’t work. ? steve ? From: [email protected] <[email protected]> On Behalf Of Andersen, Clark Isaac via groups.io
Sent: Tuesday, September 10, 2024 14:45 To: [email protected] Subject: [ucinet] LR-QAP Error Help? Is it me? Are my networks bad? #error #help ? Hey everyone! Looking forward to seeing if anyone is able to help me. ? I'm working on testing whether one network ("Who learns from whom") can be predicted by other networks ("Who trusts whom", "who recognizes whom", "Who has worked together before"). In 2 out of 3 of the college classes I've done this survey, I've gotten excellent results, great odds ratios and p-values. I'm working on a transcript to publish now.? ? But in the 3rd course, I can't repeat the process for some reason. UCINet spits out this error:? ? And gives me this printout:? ? When I try predicting Learning from just Trust, I get this bonkers odds ratio, which I don't think is accurate: ? Again, there's nothing inherently different from this course's networks than the others. In fact, this one I'm predicting will have a fairly low odds ratio, knowing what I know about these students. If this is a software bug, is there a workaround, or has anyone else seen this, or am I doing something wrong? Any insight would be great! |
||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Hey everyone! Looking forward to seeing if anyone is able to help me.
?
I'm working on testing whether one network ("Who learns from whom") can be predicted by other networks ("Who trusts whom", "who recognizes whom", "Who has worked together before"). In 2 out of 3 of the college classes I've done this survey, I've gotten excellent results, great odds ratios and p-values. I'm working on a transcript to publish now.?
?
But in the 3rd course, I can't repeat the process for some reason. UCINet spits out this error:?
?
And gives me this printout:?
?
When I try predicting Learning from just Trust, I get this bonkers odds ratio, which I don't think is accurate:
?
Again, there's nothing inherently different from this course's networks than the others. In fact, this one I'm predicting will have a fairly low odds ratio, knowing what I know about these students. If this is a software bug, is there a workaround, or has anyone else seen this, or am I doing something wrong? Any insight would be great!
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
LR-QAP Relational effect
HI,
?
Maybe I'm looking wrong, I'm not finding the help file for LR-QAP in general. Ok, you test one way or another and in the end, you can understand the best use of the resource. However, the relational effects are not easy to interpret without a help file. Somewhere in this forum, I found:
?
? Reciprocity effect
– If j likes i at T2, then predict that i will like j at T3 ? So create matrix X1 in which X1(i,j) = 1 if T2(j,i) = 1 ? Transitivity effect (closure)
– If, at T2, i likes j and j likes k, then predict that at T3 i will like k. ? Create matrix X2 where X2(i,j) = 1 if (T2(i,j) =1 and T2(j,k) = 1)* ? Preferential attachment effect
– If T2 indegree is high, predict that all nodes will want to have a tie – X3(I,j) = indegree(j) ?
Can anyone enlighten me if the logic of the resource is as described above?
?
tnx,
?
p
|
Paulo Matui
Aritra Halder