1

Supporting Information (SI) for “Rolling Off the Tongue”

Table A. Demographic Characteristics of Bilingual Survey Respondents in Study 1 and Study 2

Study 1
(N=530) / Study 2
(N=1,131)
% Native-born / 48% / 43%
% Citizen / 83% / 79%
% Mexican ancestry / 52% / 51%
% Border state resident / 56% / 55%
% Male / 49% / 51%
Median Age / 42 years old / 45 years old
Median Education / Some college / Some college
Median Family Income / $50-$59,999 / $40,000=$49,000
Median Partisanship (7-pt. scale) / Leans Democrat / Leans Democrat

Table B. Randomization Check: Predicting Treatment Assignment as a Function of Demographic Characteristics in Study 1 and Study 2

English interview =1
(Study 1) / English interview=1
(Study 2)
Native-born / -.013 (.130) / -.006 (.083)
Citizen / -.018 (.173) / -.069 (.101)
Mexican ancestry / -.030 (.140) / -.104 (.097)
Border state resident / -.048 (.138) / .024 (.095)
Male / -.130 (.118) / .042 (.077)
Age / .242 (.289) / .001 (.002)
Education / -.069 (.203) / .011 (.043)
Family Income / -.070 (.278) / -.006 (.009)
Partisanship (Democrat) / -.006 (.031) / -.001 (.020)
Constant / .199 (.285) / .093 (.201)
LnL / -330.65 / -782.40
χ2 / 2.26 / 2.90

Table C. Distributions of Demographic Characteristics by Experimental Cell (Balance Check) in Study 1 and Study 2

English interview / Spanish interview / Chi-square test
1. Native-born (%) / 48%
42% / 47%
43% / χ2(1) = .01, p = .91
χ2(1) = .32, p = .57
2. Citizen (%) / 84%
78% / 83%
80% / χ2(1) = .07, p = .79
χ2(1) = .41, p = .52
3. Mexican (%) / 51%
49% / 52%
53% / χ2(1) = .04, p = .84
χ2(1) = 1.36, p = .24
4. Border resident (%) / 56%
54% / 55%
56% / χ2(1) = .01, p = .92
χ2(1) = .24, p = .62
5. Male (%) / 47%
52% / 51%
50% / χ2(1) = .73, p = .39
χ2(1) = .54, p = .46
6. Some college (%) / 35%
32% / 37%
32% / χ2(3) = .30, p = .96
χ2(3) = .96, p = .81
7. Leans Democrat (%) / 31%
24% / 31%
23% / χ2(6) = 3.55, p = .74
χ2(6) = 5.20, p = .52
8. 42 years or older (%) / 55%
56% / 52%
58% / χ2(1) = .60, p = .44
χ2(1) = .40, p = .53
9. $50-59,999 or higher (%) / 56%
42% / 55%
43% / χ2(1) = .08, p = .78
χ2(1) = .17, p = .68

Table D. Language Treatment Wording (Study 1 and Study 2)

English Treatment
When you enrolled to participate in surveys through our organization, you reported fluency in both English and Spanish. Given your ability in these two languages, we would like for you to complete the following survey in English.
All of the questions in this survey will be in English. Please read carefully and answer each question as an English speaker. Answer each question in English to the best of your ability. This is not a test. We are simply interested in your opinions as an English speaker.
It is important that you complete the entire survey at one time. For this reason, we will record the times at which you start and finish answering the whole questionnaire.
Click “Next” to continue with the survey.
Spanish Treatment
Cuando usted se registró para participar en encuestas a través de nuestra organización, indicó que hablaba con fluidez tanto en ingles como en español. Debido a su capacidad en estos idiomas, quisieramos que complete la siguiente encuesta en español.
Todas las preguntas en esta encuesta serán en español. Por favor lea con atención y conteste cada pregunta como alguien que normalmente habla español. Conteste cada pregunta en español, lo mejor que pueda. Esto no es un examen. Estamos simplemente interesados en saber sus opiniones como alguien que habla español.
Es importante que usted complete toda la encuesta en una sola sesión. Por esta razón, anotaremos la hora en que empiece y en que termine de responder a todo el cuestionario.
Haga clic “siguiente” para continuar con la encuesta.

Table E. RT Differences Partially Mediate Language’s Impact on Knowledge

Study 1 / Study 2
U.S.
knowledge / U.S.
knowledge
English interview (Direct effect) / .080
[.034, .127] / .082
[.051, .114]
Response time (ACME) / -.005
[-.015, .003] / -.004
[-.007, -.000]
Total effect / .074
[.028, .122] / .078
[.048, .109]
% Total Effect Mediated / 7% / 5%

1

Table F. Distributions of Emotions and Political Efficacy by Interview Language in Study 1 and Study 2

  1. Anxiety
/ Not anxious at all / A little anxious / Somewhat anxious / Very anxious / Chi-square test / N
English / 60.66%
53.97% / 22.43%
27.16% / 12.13%
14.11% / 4.78%
4.76% / Χ2(3) = 6.14,
p = .105 / 272
567
Spanish / 52.33%
44.20% / 22.48%
27.07% / 18.99%
23.76% / 6.20%
4.97% / X2 (3) = 19.12
p = .000 / 258
543
  1. Anger
/ Not angry at all / A little angry / Somewhat angry / Very angry / Chi-square test / N
English / 77.61%
68.25% / 11.57%
14.64% / 7.84%
13.58% / 2.99%
3.53% / Χ2(3) = 2.51,
p = .473 / 268
567
Spanish / 74.31% / 13.04% / 7.11% / 5.53% / X2(3) = 3.30, / 253
65.44% / 16.18% / 12.87% / 5.51% / p =.350 / 544
  1. Pride
/ Not proud at all / A little proud / Somewhat proud / Very proud / Chi-square test / N
English / 12.73%
13.96% / 13.86%
21.91% / 33.33%
35.69% / 40.07%
28.45% / Χ2(3) = 0.96,
p = .811 / 267
556
Spanish / 13.71%
11.42% / 14.92%
19.15% / 35.48%
41.44% / 35.89%
27.99% / X2(3) = 4.83
p = .190 / 248
543
  1. Efficacy
/ Strongly disagree / Somewhat disagree / Somewhat agree / Strongly agree / Chi-square test
English / 15.44%
14.72% / 20.59%
20.21% / 48.53%
43.97% / 15.44%
21.10% / Χ2(3) = 0.64,
p = .887 / 272
564
Spanish / 17.83%
19.30% / 18.99%
24.26% / 48.06%
36.58% / 15.12%
19.85% / X2(3) = 9.44
p = .024 / 258
544

1

Table G. Preferred Language Inconsistently Moderates Language Effects in Study 1 and Study 2

Study 1 / U.S. political knowledge (Traditional) / American identity / Latino identity / National origin identity / Ant-Obama affect / Ideology
(Conservative) / Partsianship
(Republican)
English interview / .056
(.047) / -.068**
(.040) / -.016
(.047) / -.086**
(.027) / .018
(.044) / .020
(.036) / -.001
(.044)
Prefer English interview / .114**
(.041) / .067**
(.036) / -.041
(.042) / -.143**
(.022) / .041
(.037) / -.002
(.033) / .011
(.042)
English x
Prefer English / .030
(.060) / .117**
(.049) / .090*
(.058) / .083**
(.038) / .025
(.052) / .015
(.046) / .037
(.056)
Constant / .406**
(.032) / .707**
(.028) / .543**
(.034) / .941**
(.013) / .297**
(.030) / .460**
(.026) / .323**
(.033)
Study 2
English interview / .107
(.108) / .094**
(.029) / .058**
(.023) / -.025
(.021) / -.037
(.032) / -.010
(.026) / -.022
(.034)
Prefer English interview / .075
(.096) / .057**
(.024) / .002
(.019) / -.097**
(.020) / .040*
(.028) / -.013
(.022) / .014
(.029)
English x
Prefer English / .213*
(.134) / .032
(.034) / -.005
(.027) / .060**
(.028) / .010
(.039) / .010
(.032) / .036
(.041)
Constant / --- / .454*
(.020) / .589**
(.016) / .894**
(.015) / .426**
(.024) / .496**
(.018) / .344**
(.024)

Table H. Native-born Status Inconsistently Moderates Language Effects in Study 1 and Study 2

Study 1 / U.S. political knowledge (Traditional) / American identity / Latino identity / National origin identity / Ant-Obama affect / Ideology
(Conservative) / Partsianship
(Republican)
English interview / .064*
(.040) / -.038
(.034) / .040
(.039) / -.030
(.025) / -.017
(.034) / .001
(.031) / -.012
(.038)
Native-born / -.024
(.041) / .078**
(.035) / .019
(.040) / -.082**
(.025) / -.053*
(.036) / -.003
(.033) / -.085**
(.040)
English x
Native-born / .023
(.059) / .089**
(.047) / .002
(.056) / -.005
(.040) / .106**
(.051) / .060*
(.044) / .073*
(.055)
Constant / .490**
(.028) / .712**
(.024) / .510**
(.028) / .890**
(.016) / .348**
(.025) / .460**
(.022) / .371**
(.028)
Study 2
English interview / .136*
(.084) / .109**
(.020) / .056**
(.016) / .000
(.017) / -.028
(.024) / -.015
(.020) / -.012
(.024)
Native-born / -.020
(.092) / .116**
(.022) / .034**
(.017) / -.106**
(.020) / .031
(.027) / -.022
(.022) / -.033
(.027)
English x
Native-born / .277**
(.129) / .019
(.030) / -.001
(.025) / .037*
(.028) / -.006
(.038) /
.026
(.031) / .034
(.038)
Constant / --- / .440**
(.015) / .577**
(.012) / .871**
(.013) / .440**
(.017) / .497**
(.014) / .368**
(.018)