6684	Assessment Initial Draft	Final Draft comments
Abstract	No abstract in this draft, but this should be easy to devise from the introduction and problem statement.  However, the introduction goes into more factual detail than it needs to, which delays the introduction of the research question.  Consider some reorganization here.	The abstract is very good.
Research Question	The research orientation is clear, but it's current broad sweep of multiple questions perhaps shows too much ambition and a lack of sufficient focus.	The research objectives are sound and clearly presented.
Literature Review	The literature review is very good.  It could use another round of editing for typographical and minor syntax errors.	The literature review is substantively the same as the earlier version, except a couple paragraphs that no longer fit the changed focus have been removed (and some minor editing fixed writing issues).  Since the earlier literature was already very good, this is fine
Data & Analysis	What data is going to be used and how it will be used is unclear in the proposal.  For the U.S., the draft proposal mentions Simmons (which is not commonly used data in sociology).  The data they supply appears to include the desired variables, but it is unclear that they allow anything much more than cross tabulation tables through the OneView portal and they do not appear to provide access to download the data.  The proposal should clarify with some certainty the availability of these data.  If it is not readily available, then gaining access or identifying an alternative should be a priority.  Similarly, the proposal should provide a confident statement about the availability of the Eurostat data. It seems worthwhile to reconsider the plan to limit the research to Millennials. Any recoding or other similar work on the data is going to be the same regardless of whether it applies to all the data or just Millennials.  Keeping all the data will allow comparisons across generations (which would add considerably to the power and scope of the research) without much extra effort, without impeding analyses limited to the Millennials (which simply requires filtering).  At the very least, this will enable testing whether Millennials are indeed different from prior generations.	Overall, the discussion of the data, variables, and model is good.  Note, however, that it is not obvious that the proposed model applied to this data will permit the causal inferences being sought.  So, some adjustments might become necessary once the data analysis begins. As an aside, let me recommend against using variable names as they happen to exist in a data set (e.g., mobilpo) in papers.  It is true that scholars still do this (less, I would like to believe, than in the past), but let me suggest it is a misguided practice regardless who does it.  It main practical effect is to make work for readers who have to learn the meaning of the names.  Long ago when typesetting was expensive and statistical software limited, it made some sense, but today it seems to persist mainly because people associate it with the appearance of professionalism.  (The same goes for acronyms.)  Instead, use an obvious name built from normal words, such as political mobilization, and do the same when presenting data in tables and graphs.  Rather than declare that the project will not separately score different kinds of participation or the like, it is usually better to either omit any reference to the possibility or use more open ended language.  We cannot know in advance what we might want to do in the face of unforeseen results during the data analysis, so we want both to be and appear to be open minded and flexible.
Causal Interpretations	As presented, the hypotheses appear to be empirical guesses based on past empirical findings.  Ideally, hypotheses about empirical outcomes would represent the application of a theoretical argument, not the simple projection of past empirical findings.  What appears here seem more like informed expectations than hypotheses.  Note, in addition, that as some of the cited data are fairly old, and these phenomena change rapidly, that it may be difficult even to say that these are informed expectations rather than simple guesses.  (Note, however, that if we have no recent research on these characteristics, it is actually good news for the study, as it means it is looking at something that is unknown.) It would improve the causal interpretations, if the proposal would discuss why social media might have a positive or negative effect on political participation.   Providing an analysis of possible mechanisms linking causes and outcomes will give weight to the hypotheses.  The data proposed for the research in the proposal do not seem to match the declared intention to use data gathered over time, essentially a panel study.  Of course, one can do repeat analyses using cross-sectional data from different points in time to examine the changes in distributions and relationships, but that does not appear to be the research goal.  So, one way or another, the analytical goals or the data need adjustment so that they match. Note that regression analyses do not provide correlations, but rather regression coefficients (the kind of coefficient depending on the kind of regression).  We can crudely think of each regression coefficient as representing the unique (within the set of independent variables being used) relationship of the independent variable to the dependent variable plus a proportionate amount of each relationship to the dependent variable shared with other correlated independent variables.  That is to say that to the degree that a subset of independent variables are correlated with each other, part of the relationship they have with the dependent variable cannot be distinguished, so a regression analysis assigns that shared relationship among the independent variables to their coefficients in a manner reflecting the strength of each one's relationship to the dependent variable (it is actually trickier than that, but thinking in these terms helps in the general interpretation).	(Note that the subheadings for independent and dependent variables appear to be reversed.) Overall, the discussions of causal possibilities throughout the paper are fairly good.  They will undoubtedly become even stronger with greater experience with the literature and the data. One consideration that seems neglected here is the significance of reference groups and social networks.  A fair amount of political participation involves interactions and depends on people's connections with others.  Part of the significance of social media concerns the substitution of virtual ties for direct ones.  This has varied implications.  For example, the absence of geographical boundaries online means that users of social media are potentially launched into the affiliation conditions associated with large cities, except on an even greater scale.  (I say potentially, as many people probably restrict their online exposure to a small network based on more traditional ties.)  So, social media supports the development of social ties based on selective, specialized criteria, i.e. the creation of particularistic niches.  Social media also potentially reinforces reference group membership based on illusory network ties, as with all the people who "follow" celebrities on Twitter or "friend" them on Facebook.  The main point here is that the proposal now reads as if the respondents are individuals responding to exposure to ideas, which is true as far as it goes, but pays little attention to the reality that people form political ideas and engage in political acts largely in the context of seeing themselves as members of some social network or group.
Research Contribution	The draft proposal indirectly refers to the potential value of the project in a couple places, but it would be best to have a brief, direct discussion.	The conclusion provides a fair statement of the projects general research value, although the proposal could still benefit from  a bit more direct assessment that directly contrasts the plan with past work.
Citations & Bibliography	The citations in the text seem fine.  The bibliography needs a bit of work.  The bibliography should be sorted by the authors' last names and most standards require italics for the titles of books and journals.	The citations are fine.  The bibliography is sound but the formatting is a bit uneven.  It appears that it was produced manually (using reference software is highly recommended), and the entries were manually formatted (rather than using a MS style, which is also recommended).
Quality of Writing & Organization	The writing is good throughout and the draft has a clear organization.	Well written.
Priorities for Revising / Responsiveness to Feedback	Clarify the data to be used, better identify the alternative theoretical models that guide the research choices, and consider focusing the research topic more and expanding the range of respondents more (not limiting it to millennials).	Excellent revising work overall.
Miscellaneous Notes
Proposal		This is a well-developed proposal that should result in a good research project.
Class Overall		Exceptional work throughout the course.