ABCDEFGHIJKLMNOPQRSTUVWXYZ
1
ABC News/Washington Post poll sample dispositions
2
This spreadsheet lists sample dispositions for ABC News and ABC News/Washington Post polls as archived by ABC News from 1999-2010 and by Langer Research Associates as ABC News’ polling provider since 2010. Dispositions were produced by the field work providers for these surveys, chiefly TNS Intersearch (through June 2011) and Abt SRBI/Abt Associates (since then). Sample dispositions were not obtained for ABC News or ABC News/Washington Post polls before May 1999, and dispositions for some surveys conducted since then (mainly brief omnibus inserts) are not collected.

Langer Research Associates does not calculate response rates for ABC News/Washington Post polls, given the vagaries of doing so and what we see as the dubious value of the exercise. Researchers consistently have found that response rates are a poor indicator of data quality. (See Curtin, Presser, and Singer 2000; Keeter et al. 2000; Mariolis 2001, 2002; Merkle and Edelman 2002; Keeter et al. 2006; Groves 2006; Holbrook, Krosnick, and Pfent 2008; and Keeter 2012, as well as this 2003 summary: http://abcnews.go.com/images/pdf/responserates.pdf.) The judgments required in recording sample dispositions can produce substantial house effects in disposition reports from field work providers; one example is whether or not interviewers are instructed to apply separate codes to answering machines and voicemail messages if they are clearly identifiable as businesses or residences; another is the procedures used to code screen-outs. The further judgments required in coding and categorizing dispositions can create differences in response rates estimates. We make sample dispositions available for others to use in calculating response rates as they wish.

Computations of survey response rates should take sample designs into account. For example, to reduce non-coverage, ABC News/Washington Post polls conducted since 2005 have not scrubbed landline samples for listed business telephone numbers (see Merkle et al. 2006). This higher coverage increases noncontacts, and decreases eligibility, because it includes business listings that, in lower-coverage designs, would have been removed without being dialed. Further, from October 2008 through June 2015, ABC/Post polls employed a non-overlapping dual frame design with separate samples of cell-phone-only and landline respondents. Respondents in the cell sample were screened out if they also had a landline. Therefore, during this period (poll 1075 to 1169), estimates of eligibility in the cell phone sample should be different from the “e” of the landline sample, taking into account cases in which phone status cannot be determined for eligibility purposes (e.g., unanswered numbers, voicemails, initial hang-ups and incomplete screenings), as well as age screen-outs. In July 2015 we adopted an overlapping dual frame sample design, in which cell phone respondents are interviewed regardless of whether or not they also have a landline (see ABC’s polling methodology: http://abcnews.go.com/US/PollVault/abc-news-polling-methodology-standards/story?id=145373). Sample designs should be considered in calculating response rates using these or any sample dispositions.

Starting with Poll 1170, the cell phone only (CPO) sample is now a cell phone (CELL) sample, reflecting the overlapping dual frame. The labels for the sample have been updated for polls 1170 and beyond.

Langer Research Associates, April 2016
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100