Since its start in 2003, the American Life Panel has expanded from about 500 to over 5,000 U.S. households. The ALP recruits participants from several sources, including representative samples of U.S. consumers. Sources of participants are:
- University of Michigan Monthly Survey (MS) internet-panel cohort
- University of Michigan MS phone-panel (CATI) cohort
- National Survey Project cohort
- Snowball cohort
- Mailing experiment cohort
- Phone experiment cohort
- Vulnerable population cohort
- Respondent-driven sampling (RDS) cohort
- ALP Intergenerational Cohort
- ORC Phase 1
A recruitment_type variable indicates the origin of each household. Values for this variable are:
- 0 MS Internet
- 1 MS CATI
- 2 Snowballs
- 3 National Survey Project
- 6 Mailing Experiment
- 7 Phone Experiment
- 8 Vulnerable Population
- 9 RDS
- 10 ALP Intergenerational
- 11 ORC
The ALP also invites other (adult) household members of the sampled panel members to join, thus allowing intra-household comparisons. These panel members have a numeric identifier ending in a value greater than 1 (e.g. identifier 10017494:2). At present, approximately 17 percent of surveyed households have more than one panel member. As a result, the ALP cannot be used as a proper household-survey panel and should be considered primarily a panel of individuals. All individual respondents are assigned the household recruitment_type variable of the originally sampled household member (i.e., invited household members receive the recruitment_type value of the original sample member). Some member-referred respondents have since left the origin household and established separate households. In order to easily identify how individual respondents joined the panel, a member level recruitment type variable (recruit_type) has been constructed that indicates how individual members were added. Values for this (string) variable are:
- “ALP mailout” — mailing experiment
- “ ORC_Phase1” — random digit dials
- “added member” — referred by originally sampled household member
- “ alp_coldcall” — phone experiment
- “alp_hispanic_recruitment” — primarily RDS
- “alp_intergenerational_recruitment” — relatives referred by ALP members
- “michigan” — MS Internet
- “snowball” — Snowball
- “srg” — MS CATI
- “stanford” — National Survey Project
- “vulnerable_pop” — Vulnerable Population
Below we present more details on each source of ALP participants.
University of Michigan Internet-Panel Cohort
University of Michigan MS Internet-panel cohort respondents are those recruited among individuals age 18 years or older who had responded to the Monthly Survey of the University of Michigan's Survey Research Center (SRC). The MS is the leading consumer-sentiments survey, incorporating the long-standing Survey of Consumer Attitudes, and is used to produce the widely-used Index of Consumer Expectations. Each month, the MS interviews approximately 500 households, 300 from a list-assisted random-digit-dial (RDD) sample and 200 re-interviewed from the RDD sample surveyed six months previously. The SRC also screened MS respondents for the ALP, asking whether they would be willing to participate in a long-term research project (with approximate response categories no, certainly not, probably not, maybe, probably, yes, definitely). Respondents answering anything but no, certainly not were told that the University of Michigan was undertaking a joint project with RAND. Interviewers then asked respondents if they would object to the SRC sharing information about them with RAND so that RAND researchers could later ask them if they would be willing to actually participate in an Internet survey. Respondents who did not have Internet were told that RAND would provide them with free Internet. Respondents who initially refused were interviewed again, and told of a reward of $20 for each half-hour interview.
Through August 2008, 51 percent of the Michigan referrals agreed to be considered for the ALP, and 58 percent of these participated in at least the household information survey. That is, about 30 percent (58 percent of 51 percent) of the Michigan recruits became ALP participants. Originally, the ALP included only respondents 40 years of age and older. Since November 2006, it has included respondents 18 years of age and older.
University of Michigan Phone-Panel Cohort
University of Michigan phone-panel cohort respondents are those who originally were part of a phone panel comparing CATI (computer-assisted telephone interviewing) with Internet interviewing. After that study was completed, researchers invited them to join what became the ALP, forming the MS-CATI cohort.
National Survey Project Cohort
National Survey Project cohort respondents are former members of a panel originally recruited by researchers at Stanford University and Abt SRBI. After August 2008, the University of Michigan decided to use MS respondents for one of their own projects, so that the ALP no longer received new respondents from the University of Michigan. Instead, in the fall of 2009, ALP researchers recruited participants from the Face-to-Face Recruited Internet Survey Platform (FFRISP). The FFRISP was an NSF-funded panel conducted by researchers at Stanford University and Abt SRBI. Stanford and Abt researchers selected a representative sample of respondents who were at least 18 years old, resided in a household in the contiguous United States, and who were reportedly comfortable speaking and reading English from June to October 2008 in a multi-stage procedure based on address lists.
Stanford-panel respondents were recruited in face-to-face interviews. They were offered a laptop (worth $500) and a broadband-internet subscription, or $200 upfront and $25 per month (for 12 months) for those who already had a computer and internet access. They also paid respondents who accepted the laptop $5 for each monthly survey. Altogether, this effort recruited 1,000 respondents from a gross sample of 2,554 addresses that were not known to be ineligible. Sometimes the generosity of the incentive was a source of refusal. Often, though, initial contact resulted in an unspecified refusal, without the interviewers having a chance to explain the study. Some examples of groups of people who refused include the elderly (7% of refusals), technophobes (14% of refusals), and skeptics (4% of refusals). The Stanford panel was terminated after September 2009, but the 1000 participants were offered the opportunity to join the ALP under the same conditions (laptop, high speed internet, monetary compensation). From these 1,000, 457 agreed to join the American Life Panel.
We recruited a subset of respondents through a so-called snowball sample. The snowball cohort resulted when researchers gave respondents the opportunity to suggest friends or acquaintances who might want to participate. RAND then contacted these persons and invited them to participate. Because this "snowball" sample is not randomly selected or representative of U.S. residents, it is used mainly for pilot tests of surveys. However, they can also participate in regular surveys. No new snowball respondents have been permitted to join the ALP since May 2009.
Mailing and Phone Experiment Cohorts
We have also recruited a mailing-experiment cohort and a phone-experiment cohort as part of an experiment to test different recruitment methods. People in the mailing experiment were approached via postal mail, whereas people in the phone experiment were contacted by phone. In both cases the participants were randomly drawn from nationally representative samples.
Vulnerable Population Cohort
We expanded the American Life Panel with 2,000 panel members drawn from vulnerable groups and minorities. This addition includes a subsample of approximately 150 households for whom the interview language is Spanish. The expansion was completed by the end of 2011. We recruited these respondents from an address-based sample in ZIP code areas with high percentages of Hispanics or low-income households. We mailed potential panel members a letters (including a prepaid incentive) and made follow-up phone calls.
Respondent-Driven Sampling (RDS) Cohort
We experimented with a respondent-driven sampling (RDS) approach to sample populations through social networks with a resulting cohort of roughly 400 respondents (Heckathorn 1997; Heckathorn 2002; Heckathorn 2007). In RDS, each respondent recruits a fixed number of friends in the target population who in turn become the next generation of respondents. Once sample equilibrium has been reached, sample proportions for a given variable of interest no longer change. However, these sample proportions in equilibrium will be different from proportions in the population, because respondents with larger social networks will be overrepresented. Biases may be corrected to derive unbiased population estimates (Heckathorn 2002).
We are currently expanding the American Life Panel with panel members age 18+ drawn from an ongoing RDD recruitment using a dual frame sampling design. This means that the sample is drawn from two independent sample frames - one for landlines (60%) and one for cell phones (40%). Individuals are initially contacted through RDD, given a brief description of the American Life Panel, and are asked if they are interested in participating. If they agree to participate, they are either contacted via e-mail and then complete an initial survey online, or they are provided with a paper survey if they do not have an email address. At the end of the initial survey they are invited to become permanent ALP members. Individuals without internet access will be offered a laptop and internet subscription paid by RAND.
ALP Intergenerational Recruitment
In 2013, ALP members were invited to refer family members, potentially outside of the household, to join the panel.
The primary household respondent was invited to refer other (adult) household members to join, thus allowing intra-household comparisons. Primary respondents across all household recruitment types referred roughly 900 household members, some of whom later formed separate households. Household member referrals ended in 2014.
Heckathorn, D. (1997). "Respondent-driven sampling: a new approach to the study of hidden populations." Social Problems 44(2): 174-199.
Heckathorn, D. (2002). "Respondent-driven sampling II: deriving valid population estimates from chain-referral samples of hidden populations." Social Problems 49(1): 11-34.
Heckathorn, D. (2007). "Extensions of respondent-driven sampling: analyzing continuous variables and controlling for differential recruitment." Sociological Methodology 37(1): 151-207.