Table 1

Construct validity hypothesis tests

HypothesisVariableScore range (median)Mean rankSignificance test and p value
Level 1: Strong hypotheses from qualitative findings:
Higher scores will be associated with continued pregnancies and lower scores with pregnancies ending in abortion.Pregnancy outcome:
continued pregnancy0–12 (11)393.53Mann-Whitney U, p<0.0001
abortion0–12 (2)98.48
“Living with husband” status will be associated with higher scores, other categories associated with lower scores.Marital status/live with:
husband0–12 (11)431.83Kruskal-Wallis, p<0.0001
partner0–12 (6)283.25
not husband or partner0–12 (2)145.68
Older age will be associated with higher scores (although the full range of scores likely on all ages).Age group:
<200–12 (3)154.90Kruskal-Wallis, p<0.0001
20–240–12 (4)240.18
25–290–12 (9)307.57
30–340–12 (11)398.96
35–390–12 (11)397.37
40+0–12 (10)365.29
Higher educational status will be associated with higher scores (although full range of scores likely for all levels of educational status).Educational level:
school0–12 (6)278.92Jonckheere-Terpstra, p<0.0001
post 160–12 (7)297.97
higher/further0–12 (11)367.32
Level 2: Hypotheses from the literature, neither proved nor disproved by qualitative findings:
Black women will have lower scores.Ethnicity:
White British0–12 (10)335.59Kruskal-Wallis, p = 0.013
White other1–12 (10)331.37
Asian/Asian British1–12 (10)331.63
Black/Black British0–12 (5)246.75
Mixed/other1–12 (8)272.32
The second child (actual births) will have the highest scores, and the third-plus child lowest scores.Child order:
Continued pregnancies only:
was/will be first child1–12 (10)251.09Kruskal-Wallis, p<0.0001
was/will be second child1–12 (11)283.62
was/will be third or more child0–12 (9)197.32