The list experiment only produced higher abortion estimates than the direct questions among respondents not at all acquainted with the interviewer when using the four‐category acquaintance variable, but the difference did not reach statistical significance.
1
—
—