2016-2017 Latino Immigrant National Election Study (LINES)

<p align="center">2016-2017 Latino Immigrant National Election Study (LINES)<br />
PIs: James A. McCann and Michael Jones-Correa</p>

<p align="center"><strong>Study Details</strong></p>
<p><u>Sampling population</u></p>

<p>Foreign-born Latinos over eighteen; contact information obtained from marketing research firms.  For further details, see James A. McCann and Michael Jones-Correa, <em>Holding Fast: Resilience and Civic Engagement Among Latino Immigrants</em> (Russell Sage Foundation, 2020).</p>

<p><u>Survey design</u></p>

<p>The first survey wave was administered nationally by telephone to a representative sample (<em>N</em>=1,800).  Both cellular and landlines were called.  Nearly all interviews were in Spanish, by the respondent’s choice.</p>

<p>After the 2016 election, 576 immigrants took part in the second survey wave, which was fielded during the presidential transition period (November 8, 2016 – January 20, 2021, a 32 percent re-contact rate).  At this time, an additional fresh sample of 260 Latino immigrants was added to the study using the same sampling procedures as the first wave.</p>

<p>In the summer of 2017 (July through early-September), a third wave was conducted, with all 1,800 immigrants from the pre-election baseline survey being eligible for interviewing. In this period, 31 percent of these immigrants (<em>N</em>=554) were surveyed; this includes 321 respondents who had taken part in the second wave and 233 who had not.   In addition to these respondents, 500 fresh immigrants were sampled using the same procedures as the first wave.  In total, 2,560 immigrants took part in the 2016-17 LINES.</p>

<p><u>AAPOR response rates</u></p>

<p>RR 1 = .034 (pre-election); .020 (post-election transition); .032 (summer of 2017). </p>

<p>COOP 4 = .239 (pre-election); 253 (post-election transition); .296 (summer of 2017).</p>

<p><u>Weighting</u></p>

<p>Distributions of socio-demographic variables were compared to the American Community Survey (ACS).  In most respects, the LINES sample conformed to the ACS, though significant discrepancies were found for education, age, gender, and citizenship status.  Weighting values were calculated based on these variables through iterative proportional fitting (“raking”).  These values are given in the variable “weight.”</p>

<p><u>Support</u></p>

<p>Russell Sage Foundation; Purdue University; Cornell University; University of Pennsylvania.</p>