Click on the following links. Please note these will open in a new window.

Group Projects

1. If possible, contact your school’s office of Institutional Research and/or Registrar to get statistics about the student population.  It should not be difficult to get a series of characteristics about the population:  gender, class, race/ethnicity, major, residence (on/off campus), etc.  Assign a group of students to each statistic.  Ask them to design their own way to measure their characteristic.  Assign each group to record the characteristics for every other member in your class (you will have to provide a list of students in your class).  Have them tally up their results and decide if your class is representative of the larger student population.  Ask them to explain how this would affect the generalizability of any research done in this class to the student population.
    
2. Have each student gather data on individuals they know on campus that are outside of this class or program. Have them include the same broad indicators present in Project #1.  This will yield a sample size ten times larger than the in-class sample.  Again, tally up the results and decide if additional availability sampling approaches representativeness of the population. Have students list where/how they contacted their responses. As a group, discuss how identifying people from these unique places might change/affect the generalizability and make up of the sample. Take this moment to also discuss any issues that might relate to redundancies in data (aka if two students asked the same people).
    
3. Send teams of students on a literature search to find the best and the worst samples.  The search should be restricted to two journals, perhaps one highly ranked journal like the Child Development or the Social Service Review and one oriented to practitioners like Social Work.  Only articles based on survey research that seeks to produce generalizable findings are eligible.  Which of the researchers seeks to produce the most ethnically diverse sample?  What strategies are used by the researcher to ensure this diversity? Each student should nominate a sample with a brief justification of why it should be considered the “winner”.
                    
4. Discuss the need of diversity of respondents in social work research specifically. Have each group of students research a topic of great interest to the social work profession (such as mal-nutrition, poverty, homelessness, discrimination). Have them detail the populations most present in the research studies they read. Discuss the extent to which these studies have generalizable or ‘useful’ results. Provoke a conversation around questions similar to: “if African-Americans are almost exclusively sampled in studies around discrimination, is the research done correctly?”
    

Mini Projects

1.  A Simple Random Sample 
For this project, you and a partner will need to locate an online forum that has at least 100 job ads, rental ads, or other advertisements. Plan to draw a random sample of 25 advertisements.

1. Identify the URL on which the employment, car, house, or rental ads appear.
2. Print copies of the pages and distribute them to each member of the pair.  After this step (until Step 9), individuals in the pair should work separately.
3. Number each ad consecutively.  Note the total number of ads.
4. Using your random numbers table, determine the number of digits that you will need in your numbers to identify cases in your “population” of ads.  (For example, if there are 152 ads, you will need to use 3 digit numbers.)  Decide on a rule for selecting numbers in the table, such as every number as you move from right to left across a row and then from left to right back across the next row.  Close your eyes and pick a starting point.  Following your selection rule, write down the first 25 numbers that you encounter that you encounter that fall within the range of ad numbers (if there are 152 ads, this means you should write down the first 25 numbers between 1 and 152).
5. Circle the ads corresponding to the numbers you have written down from the random number table.
6. Record one or two characteristics of each ad.  These might be the type of job advertised, the community where a house or apartment is located, whether the asking price is listed, or the asking price itself.
7. If you used types of jobs or communities as a characteristic, review these types and, if it seems warranted, group them together in a more limited number of categories (for example, white collar and blue collar, or north side and south side).
8. Now tally up the characteristics (for example, how many housing ads listed an asking price and how many didn’t?)  Calculate an arithmetic average for the prices or salaries you recorded (add up all the prices or salaries and then divide by the total number of prices or salaries listed.)  These are your sample statistics.
9. Compare your sample statistics with those of your partner. 
10. Write a brief statement together in which you discuss the similarity or dissimilarity of your samples and whether you think your samples accurately reflect the population of ads. 

2.  A Systematic Random Sample 
For this project, you will need to secure a copy of the most recent directory of faculty members at your college or university.  If at all possible, make photocopies of the relevant pages so you can write on them.  Directories are often available in academic departments, the library, and on websites.   

1. Go through the directory and cross listings that are not specific faculty members (e.g., “President’s Office” or “Graduate Office”.
2. Count the number of remaining elements.  This is your population (all faculty members listed in the most recent directory, NOT all faculty). 
3. Designate a sampling interval in which you will end up with a sample of 25 individuals.  (This can be figured by dividing the total number of elements in the population by 25). 
4. Randomly select a starting point in the directory, preferably through use of a random number table.
5. Write down every nth name from the directory and his or her department or other affiliation (where n = the sampling interval designated in #3 above).
6. After you have written down the names and departments for each member of your sample, compile summary statements about the same in terms of gender of faculty member and department.  Write a brief statement describing the findings about the population based on the sample.
7. Contact the people selected for your sample, explain that you are doing a class exercise on sampling procedures and that you have selected them as part of your sample.  Ask them a pre-determined question of some interest or relevance (but make the questions short, easy to answer, and relevant, such as:  When did you receive your highest degree and/or from where did you receive it?).   
8. Write a brief essay about this research, in which you describe your sampling technique and your findings.

3.  Diverse Population Recruitment Strategy 
 

Ensuring diverse populations when conducting any research is fundamentally important. When working within the frame of Social Work, this need becomes even more necessary, as social workers engage with diverse populations each and every day. They need to be able to know which strategies will work to help all people, not only those who were chosen to participate in a specific set of studies.           

1.Identify a social issue present in your local or college community.
2.Collect sources, both academic and journalistic, that detail the issue, as well as mention traditional populations that are vulnerable to this situation.
3.From that list, brainstorm other groups or individuals who may be affected by this phenomenon.
4.Produce a new strategy (or modify an existing strategy) in order to craft an inclusive, but binding strategy for recruiting a diverse group of individuals to produce in the study you would like to conduct. Think about including:

a.Different ways to advertising your research study
b.Potential incentives
c.Different sampling techniques as well as their associated weaknesses and risks.

4.  Snowball Sampling

Select a topic for investigation that focuses on a student characteristic or behavior that is relatively uncommon, unorganized, and not readily apparent.  Some possibilities are strict vegetarians, guitar players, recreational long-distance campers, and wilderness campers (try to avoid sensitive topics).   

1. Compose two questions to ask people in this population.  For example:  How did you get involved in this activity?  What do you like and dislike about it?
2. Identify a student in this population, using personal knowledge, a newspaper article, or a friend’s reference. 
3. Ask this person if you can ask them a few questions about the activity for a class project.  If they agree, ask them the two questions and then ask them who else they could suggest as participants in this activity.  If they decline, just ask them if they can recommend anyone else you could talk to about the activity.
4. Repeat this process, asking the new contacts for names of people whom they would recommend.
5. You can stop after you have three sets of names:  from the initial contact, from the persons he or she recommended, and then from the additional contacts these persons recommended.
6. Write up a description and critique of this snowball sampling process.  How many persons were in your final list? What biases, if any seemed to be apparent in the recommendations of people who were recommended by other people who were recommended by other people, and so on?  How many stages do you think there would have to be to your snowballing to develop a complete list of all students involved in this activity on your campus?

5.  Respondent-driven sampling

In groups of 4 or 5, identify a non-profit or public sector organization in your local community that at least one of you has familiarity with. Find contact information and reach out to them, asking for an interview with someone who works on developing and managing programs.

            During the interview, ask the manager how s/he goes about choosing/developing programs. Specifically ask if s/he ever uses research or relevant literature to help them make their decisions. If yes, ask for a specific recent problem that literature helped them solve. Review the literature on this subject and write a few paragraphs detailing whether or not the manager seems to have correctly identified the relevant information (with regard to generalizability) when making their decision.

            If the manager does not currently use the literature to help inform his/her program decisions, ask them to identify a program related question they’ve recently been trying to solve without much resolution. Then follow similar steps as stated in the previous experiment, except that, in this case, you should present your conclusions based on a review of the relevant research to help the decision maker make his/her decision. Remember to look into research design, sampling methods, and the broader issue of generalizability.