Science topic

Group Analysis - Science topic

Explore the latest questions and answers in Group Analysis, and find Group Analysis experts.
Questions related to Group Analysis
  • asked a question related to Group Analysis
Question
2 answers
Hi,
I have a series of quantitative MRI images (T1, T2, and myelin water fraction maps) for both patients and controls. I want to do a whole brain vowel wise analysis but am not sure what would be the best software to use. SPM12 is what comes up a lot on google but it seems like its built for functional images. My current idea would be to just brute force it with a MATLAB script to calculate Z-scores. Are there any programs that is already made for this type of analysis?
Thanks
Relevant answer
Answer
thanks i'll look into those programs. What would be the most effective way to co-register patients and controls amongst each other so I can average the images?
  • asked a question related to Group Analysis
Question
7 answers
Dear colleagues,
I am testing whether a variable mediates or moderates a relationship across two groups in a SEM. All variables are continuous and latent.
Is there a way I can model latent interactions in SEM and still being able to get a LR and overall goodness of fit indicators so that I can compare the moderation model with the mediation one?
Anyone ever attempted with the med4way package?
Thanks in advance
Relevant answer
Answer
There are three main approaches for doing latent interactions.
  1. Product indicator approaches
  2. Distributional analytic approaches
  3. Instrumental variable approaches
Approaches 1 and 3 can be readily implemented using Stata's SEM. Approach 2 is not supported by Stata. Mplus supports LMS as Mehmet says, which falls into this category. But note that distributional analytic approaches are sensitive to normality assumption of the latent variable. This assumption is untestable and therefore I personally always prefer to go for product indicators.
I cover all these approaches on my doctoral level course. You can find explanations here: https://www.youtube.com/watch?v=Ibv10E_WRYM&list=PL6tc6IBlZmOU_R1Ezb1WShAmLAuCVwF4S
(Note that at the time of writing I have not uploaded the interaction videos but will do that in the near future.)
  • asked a question related to Group Analysis
Question
3 answers
Hi peers,
I used AMOS to test the moderating effects of demographic variables (gender, age, marital status, education, and income). However, the results of multiple group analysis and interaction moderation are largely different. (n=302)
In particular, when using MGA for testing moderating effect, marital status significantly moderates the effect of the independent variable on the dependent variable,
Single Married
Estimate P Estimate P z-score
ITUT99 <--- EFCT99 0.335 0.001 -0.033 0.686 -2.802***
ITUT99 <--- PFCT99 0.261 0.008 0.656 0.000 2.894***
but when using the interaction method, these interaction effects are non-significant. (I standardized values of the independent variables and demographic variables, interaction = Zindependent*Zmaritalstatus)
Estimate S.E. C.R. P
ITUT99 <--- PROxMAR .221 .205 1.076 .282
ITUT99 <--- EMOxMAR -.208 .206 -1.012 .312
I am very grateful if you can help me explain why there is a big difference in results between the two methods. Are there any ways to improve the results in the interaction effect?
Relevant answer
Answer
Hoang T.P.M. Le Engagement is more general than moderation. The function of the two variables involved in the interaction is distinguished by moderation. So, when we state X and Z interact in their impacts on an outcome variable Y, there is no actual differentiation between X's and Z's roles.
  • asked a question related to Group Analysis
Question
3 answers
I have purchase data from supermarkets that include a variable for weighting. This weighting is supposed to represent how representative a household is.
I want to separate my data into two very unequal groups and aggregate the values into months. Can or should I use the weighting here?
Relevant answer
Answer
I find it a little difficult to answer this as you have not spelled out your aims.
I would keep the detailed data (presumably repeated measures over occasions on households?) and then include a set of dummies to identify the month and a dummy to identify the two groups (using between and within random effects for households and occasions). An interaction between the dummies would then allow an assessment of in which months the dummied group has a different mean. In this form I would then do a weighted and unweighted analysis as a sensitivity test to see what is the scale of the difference. Of course the weighted and unweighted analyses represent a different target of inference. By using random effects you could also include such variables as type of household if this was of interest. Using a mixed model, you could include an underlying 'smooth' (such as a spline) to capture the underlying time trend in the purchasing behaviour.
  • asked a question related to Group Analysis
Question
4 answers
After normalization and summarization of microarray raw data,
when I carried exploratory grouping analysis.
The samples from the two comparison groups don't cluster together, should I exclude the samples that cluster in the wrong group?
or should I continue the analysis to determine the differentially expressed genes without excluding any array?
Relevant answer
Answer
Not only normalizing your data, you need to do all the necessary preparations on your data such as selection for training and validation before running the exploratory analysis. Go through your steps carefully,
You may not necessarily have a clear grouping or cluster due to overlap, in this case you should be able to explain such scenario based on your samples or experimental results.
  • asked a question related to Group Analysis
Question
4 answers
I have performed a single treatment (25 micro gram) of a bacterial virulence factor in presence of multiple inhibitors to see the effect in single cell line. Untreated cells, H2O2, virulence factor alone and inhibitors alone are also considered as different controls. To compare different treatments with each-other I have performed One-way ANOWA test. Is this accurate or I'm in wrong path.
Relevant answer
Answer
It seems you made a couple of replicates (take cells, add the virulence factor, split the cells, add a different inhibitor to each well, incubate, measure -- so similar), and each series is normalized to the fluorescence reading of the control sample in this series,presumably because each time you get quite different readings between the series.
This kind of normalization not state of the art. One of its drawbacks is that you use some information twice and don't adjust the degrees of freedom accordingly and that a "bad" value in the control will spoil the entire series of measurements.
You shoud instead use a hierarchical model with a random intercept (per series). If this is not feasable you should at least use the average across all the readings within each series to normalize against.
Note: it is likely that you should analyze the log intensities, particularily when the range ob measured values is relatively large compared to the lowest (or even to the average) values.
Note: You need a correction for multiple testing. If you compare every inhibitor against the control, you should use Dunnett's procedure. If you want to do all-pairwise comparisons, you should use Tukey's HSD.
  • asked a question related to Group Analysis
Question
5 answers
Consider I have 3 groups: 1, 2 and 3. In each group I have 3 species. 1A, 1B, 1C; 2D, 2E, 2F and 3G, 3H, 3I. It is possible to predict where groups occur given some predictors. However, species in group 2 are more "generalistic" and have the tendency to occur in the other groups. While I could weigh group 1, 2 and 3 lets say 2/1/2, I actually want to weigh all species 2 times more except for 2D and 2F, which I only want to weigh 1 time. Lets say I predict group 1 at a specific site, and at this site the species composition is 1A, 1B, 2D this prediction is for 2/3 correct, but I do not see the occurrence of species 2D occurring in group 1 as very problematic if the weights of the species in this group was 2/2/1 than the model would be 4/5 correct. Is there any way to perform this with random forest models?
Relevant answer
Answer
Roberto Vega after further pondering on this "issue", this "solution" only impacts the location of split-points not the classification error. Or, am I wrong?
  • asked a question related to Group Analysis
Question
9 answers
The whole data was collected on Likert scale. I am using Prof. Gaston's plspm package in R for SEM modeling. I think different age groups must have differences. But, I can't test differences of more than than two groups at structural level. My question is, if I divide the data into various age based subgroups and prepare SEM models separately for each subgroup. Is it meaningful? How to justify the models are significantly different? I was not able to perform ANOVA test to check the difference among models. What should I use? Please guide me Thanks
Relevant answer
Answer
Smart PLS has excellent capabilities for multigroup PLS-SEM analysis. You can for instance access to statistical testing for path differences between groups. This software is not open source but you can freely download and use it for a free 30 day trial (maybe it could help)
  • asked a question related to Group Analysis
Question
4 answers
Dear Concern,
Kindly guide me regarding the step by step process of data aggregation in Mplus/SPSS. (Pls answer as per your expertise in any of the mentioned statistical software).
I have collected data from two sources; team members and team leaders. During the survey, leaders were given questionnaires for variables, and one among them was the team-level variable (group norm). Regarding this, I am not sure about its aggregation and further hypotheses testing.
It would be kind of you if your understanding of this query (data aggregation) can help me proceed with my data analysis and interpretation.
Thanks in anticipation!
Regards,
Relevant answer
Answer
Thanks, respectable(s); Amlan Haque and Kelvyn Jones!
  • asked a question related to Group Analysis
Question
1 answer
In block design task fMRI, should the response rate and reaction time at which subjects press the button for stimuli in the MRI scan process be treated as a covariate in group comparison? Most of the papers didn't treat it with a covariate. How is it right?
Relevant answer
Answer
I think those factors are accounted for at the level of the individual participant, right? Unless one is interested in looking at effects of mean RTs at the group level (e.g. whether activity in some decision-making region scales as function of how quickly participants respond across participants), predictors for button press are included when conducting a GLM at the individual level. That way, the contributions of things like button press, RT are already "take out of"/accounted for in brain activity associated with a certain condition of interest (e.g. viewing faces) that is carried over to the group level.
Hope this helps!
Mrinmayi
  • asked a question related to Group Analysis
Question
9 answers
Dear all,
I've got a question about two-way ANOVA in SPM, more specifically about which 1st-level contrasts to include in the 2nd-level ANOVA. It seems to be a rather basic question but I've seen different approaches out there...
In my case, I want to study interactions and main effects in a within-subject experiment using a 2x2 factorial design. Each subject performed the same experimental session several times (repeated measures).
It seems to me that there are in the field 2 different approaches to perform it:
1- Creating 2x2 factorial contrasts (conjunction, main effects, and interaction) in the 1st-level analysis (in SPM: fMRI model specification --> Factorial design --> New Factor, and specify the two factors) and then include each of these contrast images across subjects in the 2nd-level analysis (Factorial design specification --> Design --> One-sample t-test).
This procedure finds endorsement in https://en.wikibooks.org/wiki/SPM/Group_Analysis and SPM manual.
2- Creating simple contrasts in the 1st-level analysis (for example, task minus baseline, for every session) (in SPM: fMRI model specification --> Data & Design...) and then include these contrast images in the a Full factorial analysis (Factorial design specification).
Could you please help me to understand (1) which approach is more recommended or (2) under which conditions one or another should be used? (3) If one of them shouldn't be used, could you help me to understand why? (4) Are the approaches similar under specific conditions?
Thanks in advance!
Gustavo
Relevant answer
Answer
Let's say you have two factors, letter (A and B) and number (1 and 2) in your 2x2 factorial design. So, as described above, you've created, for each subject, contrasts A1, A2, B1, and B2 (condition minus baseline).
Then, in ImCalc you can easily compute A1 + A2 - B1 - B2 as the main effect of letter, A1 - A2 + B1 - B2 as the main effect of number, and A1 - A2 - B1 + B2 as the interaction.
I hope it helps!
Gustavo
  • asked a question related to Group Analysis
Question
5 answers
Hello,
I have a data set that can be separated into two groups and thus, I ran a multi group analysis in SmartPLS in order to compare if there are significant differences in the beta-values for the two groups.
To get the model fit indices as well as the specific indirect effects and the f² values for each group (which are not shown in the multi group analysis), after conducting the multi group analysis, I split the sample and ran the Consistent PLS Bootstrapping for each group.
Now I have the problem that the beta-values for the groups are different in the single analysis and in the multi group analysis. They differ substantially. Does anyone know why and how I can get the values I am looking for for the multi group analysis?
Thank you in advance!
Relevant answer
Answer
But is it normal to have difference in significance in the two cases? i.e. in the MGA and in the individual samples?
  • asked a question related to Group Analysis
Question
4 answers
Hi everyone,
I'm trying to analyze an existing ecoinvent dataset in Simapro and would like to present results per function (eg; transport, energy, category X, category Y, etc), to see contribution per category. However, I'm unable to use the "Analysis of Groups" features in Simapro, since the processes are looped data (grey edge around the process). Is there any way to do an analysis of groups for looped data?
Relevant answer
Answer
Hi, don't know if you're still having trouble with this but if you use system processes rather than unit you should be able to do group analysis. If there is no system option you can create a new one by making a new material/process and describing it as a system process then add whatever it is you want to the inputs and make the output the same.
  • asked a question related to Group Analysis
Question
3 answers
I have the following issue with a paper, which I think is very common:
We are studying the effect of a certain disease on white matter microstructure. We want to know if there is an effect of this disease in both (or either) males and females. We have run a whole group analysis, and after that split the group by sex to test the association in each sex separately. Our results are as follows: there is no effect in the whole group, there is an effect in males but not in females.
However, as pointed out by one of our reviewers, the correct way of testing sex effect is really to do an interaction analyses, and only split by group if the interaction term is significant. In our case there is no significant interaction term. Thus, if we choose this option, we have no effect in the whole group and no interaction effect. However, if we choose such an approach, we "miss out" on the effect of the disease on white matter in the males.
I guess this boils down to what the question really is. Do we really need to test if males and females differ significantly from each other (which is what an interaction term tests) to know if disease X has an effect on the brain in either sex? Would it not also be accurate to just split and report that we find an effect in males, but not females, and add that we cannot draw a conclusion about the females (as they might have been to low N) and not about males vs females either?
I am curious what your views are on interaction vs splitting groups.
Relevant answer
Answer
Hi!! Everytime I review a paper where the results can be split in "male & female" results, I ask the researchers to explain, for example, if other categories of analysis, such as age, ethinic background etc, would be applied as well to do some more splitting in order to perceive any kind of significant results that may be hide in a dicotomic gender split.
In my opinion, you have two options:
1) You can split and report that you find an effect in males, but not in females, and add that you cannot draw a conclusion about the females YET or
2) You can add to the analysis another categories if relevant (as age, social background, etc) and than compare each intersectional categories (example: 30 years male x 30 years female; 23 years caucasian male x 23 years caucasian female, etc) in order to explain that if it is really a gender difference that justify the different results between the groups, not others factors.
I suggest you to read "Better science with sex and gender: Facilitating the use of a sex and gender-based analysis in health research", by Joy L Johnson, Lorraine Greaves and Robin Repta.
Hope it helps. :)
  • asked a question related to Group Analysis
Question
4 answers
Dear coleagues!
I have data from three groups, two patients per group and the analysys repeated three times...
There is no normal distribution.
I was wondering how can I tell GraphPad that two patients belong to one group, the other two to another... so I can compare group to group....
Relevant answer
Answer
I'm just afraid, with just 2 samples, I'm not sure whether its statistically even valid or not.
Good luck
  • asked a question related to Group Analysis
Question
5 answers
I am running a model examining the relationship between preservice teachers' commitment, their learning, and job satisfaction. I have developed a structural model after conducting exploratory and confirmatory factor analysis (split half method). However, when conducting group analysis using whether student teachers are planning to take a high stake test, I first tested metric invariance by operating on the measurement model and found that the factor structures, means, and variances of the error terms vary. As the software provides posthoc information about locating invariances, I imposed equality constraints on all the parameters that are suggested invariant in the measurement model and began to test group invariances of the structural model. However, the model does not converge at all. Most research studies using SEM and group invariance analysis stop at the measurement model, and among the limited studies testing path eqivaluence, the prerequirement of the equivalence of measurement models is satisfied. What are some suggestions of you if you were in my shoes? I would greatly appreciate?.
Relevant answer
Answer
I detect a different issue with your approach: The split-half method is considered rather flawed when it comes to CFA. Instead, you should opt for a completely independent sample. Furthermore, scalar invariance is the level of invariance which allows you to compare means across groups. You need at least n=200 in both your groups to get realiable measures. As the RMSEA is influenced by small sample sizes, I would deem changes in the CFI <0.02 and non-significant chi-square differences from one invariance level to the next as sufficient.
I would be happy to provide sources if you need them.
  • asked a question related to Group Analysis
Question
1 answer
Hi everyone.
I am using the following pipeline to analyze my fMRI data in a multi-subject multi-session setup.
1. Pre-processing data (brain extraction, segmentation, motion correction, regressing CSF out, normalization to MNI space)
2. GLM on single session level
3. within subject fixed-effect model
4. Within group analysis
5. Between group contrast.
I use the pre-processed images in MNI space to find beta-values of my estimators.
To apply GLM, I am using FSL. After I get the single-session level (first-level analysis), I cannot proceed to the 3rd step, since FSL requires the registration files. I cannot understand why registration files are required for higher level analysis when all of the images are already in MNI space.
Relevant answer
Hi Ehsan Kakaei . FEAT first-level analysis only apply the registration to the main structural (highres) and to the example_func files; thus, all other resulting files, such as cope1 used in the higher-level analysis, are not in MNI space. Thats why mat files are needed.
  • asked a question related to Group Analysis
Question
6 answers
Hi everyone!
I have to construct a design for a GLM analysis using FEAT (FSL)
Here are the data and the experiment we did:
There are those 24 subjects, that can be grouped into 2 groups depending on their behavioural responses. These subjects while in the scanner are shown a set of information over the 6 runs for a total of 120 different stimuli/trials. Each stimulus can be classified into 4 groups: High-High, Low-Low, Low-High, High-Low. The subject responds with a score 1-4 to each stimulus - this is our behavioral DV. What I would need to find is a) are there differences and where between the two behavioral groups? (both anatomical and functional) 2) are there differences and where depending on the 4 stimuli presented in the entire sample? 3) are there differences and where depending on the 4 stimuli presented in each behavioral group? 4) how does all of these above link to the behavioral DV?
These are my questions about the GLM analysis:
1) Would it be better to analyze each run of each subject and then do an analysis on all the run? After this the group analysis. Or to concatenate the runs and then do the within subject analysis and then the group one?
2) Should I include in my GLM only the 4 conditions or all the stimuli?
A good approach would be to do a 2x2 factor ANOVA analysis, but I still have some doubts on it.
What would you suggest me to do?
Thank you,
Carlotta.
Relevant answer
Answer
In attachment a better SAS program and the same analysis using R. Theu both yield the same results (of course); only the 2x2 interaction is sign. in this example.
  • asked a question related to Group Analysis
Question
2 answers
I have 3 different groups and want to run a between group analysis. Group 1 = n of 20. Group 2 = n of 30 and final group n of 200. Group 1 and 2 are older and heavier than group 3. I want to match the groups so that I can run statistics to look for between group differences in cardiac variables, but am unsure of the appropriate way to do this. Can anyone help? Many thanks in advance
Relevant answer
Answer
with non-parametric statistical techniques SPSS
  • asked a question related to Group Analysis
Question
8 answers
Dear Professors and Researchers,
I am doing a SEM modeling using SamrtPLS. In my reserach I have categorized the age group into three groups such as Young adults, Middle aged adults and older adults.
While doing multi-group analysis for gender, I did it with two groups (male & Female) and got the result. In the same way I want to test my model with age group (Young adults, Middle aged adults and older adults). Smart PLS doesn't allow me to choose more than two groups. How can I test the model with difference of age group among three categories.
Looking for your guidance.
Thanks in advance.
Relevant answer
Answer
Hi Dinesh,
Answering your question with the assumption that gender and age groups are moderators in your study. Yes PLS not allowing to compare 3 groups at once. What you can do is create pairs at a time and compare as young vs Middle, young vs older, middle vs older.
But make sure you meet the condition/assumption of MGA:
1. similar group sizes for each group because path coefficients depend upon sample size. If sub group sizes are not equal you still can create pairs as : Yong vs older (taking middle and older together) or vise versa by looking at the sample size of each sub set. Size need not to be exactly same but more or less similar size.
2. whether your moderator is considered for the whole model or for a specific path: if its for a specific path, you have to use interaction method. If it is for the whole model: MGA can be used
3. to use MGA, your moderator needs to be a categorical one (as you correctly taken)
4. Some references say data should not be too non-normal (below reference's page 487), though PLS has no normality requirement. So to avoid possible future questioning, try to see data normality also.
Find and read this reference which gives examples write ups:
V. E. Vinzi, W. W. Chin, J. Henseler, H. Wang, Handbook of partial least squares: Concepts, methods and applications. (Springer Science & Business Media, 2010).
All the best!
  • asked a question related to Group Analysis
Question
5 answers
Hi there,
I received extracted whole RNA from TIPS patient serum and negative healthy controls. All samples contained spiked-in SV40 miRNA already as a reference gene. I performed miRNA reverse transcription and analyzed expression of my miRNA of interest during Real Time-PCR. However, it turned out that expression of SV40 differed tremendously between TIPS patients and healthy controls although it should be nearly the same in the end. Because I got this samples from a neighbour group for analysis I assume different handling of those samples e.g. prolonged freeze-thawing of one set of samples. As this situation complicates a comparision of miRNA expression between TIPS and healthy controls I wondered whether there is a reference miRNA in serum which I could use for normalization instead ? Did anyone quantify circulating miRNA from human serum before using an endogenous miRNA and not a spiked-in miRNA?
I am looking forward to your experienced responses.
Thanks for helping me out.
Relevant answer
Answer
Hello Milad,
I really appreciate your help.
  • asked a question related to Group Analysis
Question
2 answers
In group analysis, this is a major concern. If we have different channels interpolated for different subjects, how can we have a concensus measure?
Relevant answer
Answer
The bad: Interpolating will create false or inflated connectivity to interpolate channels, and then calculate, e.g. coherence between the interpolated signal and the signals it was interpolated from. Systematic differences between groups will invalidate inference.
The good: The shared portion of the signal after interpolation might be about the same as if you had a clean signal and did not need to interpolate. Since you most likely interpolate from adjacent channels, the channels still pick up much of the same underlying signal to begin with, due to the spread of the electric field.
I am not aware precisely how interpolation affects connectivity compared to just having "clean" signals picking up the same sources, and not aware of any papers reporting the difference (if anybody knows, please post). I guess you could try to simulate a signal and test how interpolating channels affect connectivity.
If you do not already know of these two articles, I strongly recommend them as a reference on connectivity:
Best regards
  • asked a question related to Group Analysis
Question
1 answer
Can anyone tell me what is the difference between hierarchical bayesian modeling and multi group analysis in terms of the advantages and assumptions of using each one?
Relevant answer
Answer
Hi Khaled,
your question is very (very) broad. So I will try to answer the question on the top level: Bayesian statistics give you the information you want to have, and that classical significance testing does not provide, i.e. how confident can you be that a hypothesis is true/false. This information is conceptualized in at least two measures in the Bayesian analyses, Bayes Factors, and credible intervals for parameter estimates. Which give you information about the "odds", that your effect is found given, that there is no (or another hypothesized) effect, and whether the estimated range of plausible effects (which are considered as being likely in resulting the your observed outcomes) include the hypothesized value, e.g. a mean difference of zero. But Adrian has a point there, when it comes to implementation, you have to learn modeling before :), and you need computation power. But in the end, you have a strong control, on what you assume about the populations (hierarchically), their assumed means and distributions, which in turn enables you for implementing your hypothesis in a sound way. And moreover, another nice, although frustrating, feature of Bayesian models is, if they do not converge (in terms of distributions, not in terms of maximizing fit), then this hints you towards the possibility, that the model assumptions you made are wrong. And then you adjust them until the model finally converges. From which you can learn something of course. I am not sure, but I think, you won't get this kind of "plausibility feeling" from classical ways of significance testing.
Hope this helps for now,
Best, René
  • asked a question related to Group Analysis
Question
3 answers
Dear Claire,
Congratulations on your article about the neuroscience of the SU published in group analysis. Connecting the SU concept with the brain is an original idea.
I would like to draw your attention that Earl Hopper and I, in our introduction to the 3rd volume of our series of books about the SU, published a few months ago with Karnac, wrote about the telatikn between the mind and the SU. In addition, you might find interest in our developing ideas about this concept, and our clear definition of it, described in the introductions of the three volumes that have been published so far.
Best regards
Haim
Relevant answer
Answer
I'd compare these finds to the results in Jeffrey M. Schwartz and Sharon Begley, The Mind & the Brain: Neuroplasticity and the Power of Mental Force (Harper Perennial, 2003). You may find these results to collectively constitute an interesting conversation partner.
  • asked a question related to Group Analysis
Question
4 answers
Hello everybody,
does anyone know an easy/possible function/way/method/package to cluster spatial objects (polygons in my case, a grid of 500mx500m cells) with an equal number of objects per cluster using QGIS or R?
I'm looking for something like ArcGIS's Grouping analysis tool (http://pro.arcgis.com/en/pro-app/tool-reference/spatial-statistics/how-grouping-analysis-works.htm).
I'm attaching an example of what I mean but, in this example, the number of cells per cluster is not the same....
Regards
Relevant answer
Answer
Dear Maurizio,
Another solution could be using a fuzzy clustering approach (e. g. cluster::fanny). Fuzzy clustering provides you with a membership degree of all clusters. This is accessible as the matrix membership in the fanny-object. You can sort the membership vectors for all clusters and cut them at the required count.
Best wishes,
Oliver
  • asked a question related to Group Analysis
Question
5 answers
I have a small group all of whom responded positively to a questionnaire about perceptions of improvement since starting treatment. Being all positive (1-8 improvement on a 10 point likert scale prior till now on ratings 1-10) it should be possible to see a significance using small groups analysis. But which? I don't have access to SPSS.
Relevant answer
Answer
Helen,
It is not clear what you are trying to test, or what question you are trying to answer.
You have surveyed your subjects once or more than once?  I don't understand what "prior till now on ratings 1-10" is supposed to convey.
  • asked a question related to Group Analysis
Question
1 answer
Hello 
I have been trying MELODIC group analysis and get error after completing and didn't get final group ICA:
child process exited abnormally
> while executing
> "fsl:exec "${FSLDIR}/bin/feat ${fsfroot}.fsf -D $FD -I $session -prestats" -b $howlong -h $ID -N feat2_pre -l logs "
> invoked from within
> "if { $done_something == 0 } {
>
> if { ! $fmri(inmelodic) } {
> if { $fmri(level) == 1 } {
> for { set session 1 } { $session <= $fmri(mult..."
> (file "/usr/local/fsl/bin/feat" line 387)
 
 
When the MELODIC start, then I get this message
Feat main script
/bin/cp /tmp/feat_0y46We.fsf design.fsf
mkdir .files;cp /usr/share/fsl/5.0/doc/fsl.css .files;cp -r
/usr/share/fsl/5.0/doc/images .files/images
Please let me know how to reslove this problem.
Thanks and regards
Sadhana
Relevant answer
Answer
You can try to reinstall the software and redefine working path in terminal.
If the program works well, there are online tutorials very helpful, like this:
Best regards
Riccardo
  • asked a question related to Group Analysis
Question
1 answer
Hi 
I'm performing a SEM model were I want to add control variables.
I would like to control for gender but I have only 124 men but 366 women. Is there too much difference between the two groups to use it as a control variable? And if is significant can I afterwards do a multi group analysis?
(I am using Amos) 
Thank You in advance. 
Relevant answer
Answer
Dear, In SEM, if you are using AMOS, which has the data related assumptions (namely, minimum data size, and normality). In AMOS, you need to fulfill the condition of the data for each group before running analysis. There are different criteria (ratio with the number of items/observed variables, you may find some relevant literature for minimum sample size). However in PLS based SEM i.e. using SmartPLS, there is no data related assumptions (i.e., minimum data sample and normality assumption).
In nutshell, for PLS based analysis using SmartPLS you can safely run the analysis, however, for CB based SEM you need to fulfill the data assumptions.
  • asked a question related to Group Analysis
Question
6 answers
I have performed VBM analysis on my dataset consisting of two groups and estimate the minimal cluster size using REST AlphaSim. For the calculation of the whole brain cluster size, I used the smoothness (FWHM = 4mm) default, rmm was set at 5 and I used p-threshold of .05. As mask I defined the brain mask and 10000 iterations from the group analysis. This yields a minimal cluster size of 696 for a corrected alpha value of .05.
Does it mean that cluster size we get, is FWE corrected? If I want to run FDR correction, then I have to set p threshold of 0.01.
Relevant answer
Answer
Thanks. I know this. I am using only gray matter mask only, not whole brain mask.
  • asked a question related to Group Analysis
Question
3 answers
Any advise on the use of PhotoVoice in participatory evaluation? There are some serious limitations of this approach --- how can I overcome those?
Relevant answer
Answer
Hello Gorkey,
This paper might be of interest:
Kong, T. M., Kellner, K., Austin, D. E., Els, Y., & Orr, B. J. (2015). Enhancing participatory evaluation of land management through photo elicitation and photovoice. Society & Natural Resources, 28(2), 212-229.
and possibly these:
Dennis, S. F., Gaulocher, S., Carpiano, R. M., & Brown, D. (2009). Participatory photo mapping (PPM): Exploring an integrated method for health and place research with young people. Health & place, 15(2), 466-473.
Wang. Photovoice
Hurworth (2003) Photo-Interviewing for research
Very best wishes,
Mary
  • asked a question related to Group Analysis
Question
5 answers
If you have two groups (i.e. treated versus untreated) you would like to compare with a parametric analysis, should you determine if the dependent variable is normally distributed across all samples or is the normal distribution assumption apply to the distribution within groups and not between groups?  It seems like the normality should be determined within groups because each sample set is theoretically from an independent population (one population being treated and one population being untreated).  Any thoughts from the biostatistics community?  
Relevant answer
Answer
You are correct, the assumption for things like the t-test is that the residuals are normally distributed. If the two populations differ in means, then the distribution of the combined groups might be bi-modal (depending on how different the means are) even if both are normally distributed on their own.
  • asked a question related to Group Analysis
Question
2 answers
How can i determine optimal system(one dimensional subalgebra) using maple program in lie group analysis. Can you suggest a maple program?
Relevant answer
Answer
i read this article but i did'nt find download link.
  • asked a question related to Group Analysis
Question
4 answers
I am working on two groups. Group A has sample size of 37 and B has a sample size of 29. Can ANOVA or t-tests  affect the results due to non-equivalent sample size? How can I randomly shrink the size of subjects in Group A to 29? Is this the correct approach to Data Cleaning/Analysis? 
Another question, In case of missing values due to drop out of subjects or non-response which approach is best? Multiple imputation that will randomly fill the missing values or deletion of subject(non-respondent) and the relevant data? I am using SPSS tool for cleaning and analysis.
Relevant answer
Answer
For missing value, for your problem, imputation doesn't help.  If you have other variables, you may check the homogeneity of distributions of other variables by missing and non missing, and discuss your finding    
  • asked a question related to Group Analysis
Question
8 answers
I have successfully synthesized my nanocomposites with different concentrations. For functional group analysis i have done with IR spectroscopy. I have observed Hydroxyl group, H2O group and CO2 group attached in the first sample(Pure). As i made first composite and increased the concentration of the composites these groups are vanished in IR spectroscopy. So what are the possible reasons for this?
Relevant answer
Answer
I'm agree with Dr Malakhov the explanation could be on your reacting conditions (temperature, irradiation..) or even in your reacting mechanism. For example at room temperature reaction you can check the article of Qin et al, Graphene-wrapped WO3 nanoparticles with improved performances in electrical conductivity and gas sensing properties. J. Mater. Chem., 2011, 21, 17167-17174; where they claim the formation of O2 by water reduction on WO3, that could help to the desorption by displacement. But unfortunately their IR shows the opposite of your case. Just keep in main both; conditions or mechanism which let the desorption.
  • asked a question related to Group Analysis
Question
1 answer
In order to understand group sizes of social animals, I am looking for within-species raw data of group sizes (or colony, hive size) of social insects or social mammals. If you have unpublished raw data you would like to share or know published data, please get in touch with me.
After searching through the literature and web extensively I hope to find data this way.
Thanks a lot.
Michael
 
Relevant answer
Answer
Dear Dr Michael
Will you please specify your need. 
For what purpose you are in need of that data.
I can suggest you, better you can approach any entomologist who are continuing their research on foraging or behavioural aspects of honeybees.
GOOD LUCK