The valid response rate varied by question from 83.8% to 100% for
the fifteen questions. However only 5.1% of respondents answered Q9
about reservation services and 37.8% answered Q14 concerning the
lounge. Thismay be becausemany respondents did not use the reservation
services or lounge (for most airlines the lounge is reserved for premium
class or higher-tier FFP members, whereas the largest portion of
the sample travelled in Economy Class). Consequently, Q9 and Q14
were excluded from further statistical analyses because the proportion
of valid responseswas considered too small to represent the target population,
possibly leading to biased results.
In general, the frequency distributions tended to be skewed towards
the higher ends of the 5-point scales. The skewed distributions are
reflected by the high median scores. The results reflect the possibility
of a high level of acquiescence response bias, evidenced by the tendency
of themajority (over 50%) of the respondents to endorse, agreewith, or
provide positive responses (median scores of 3 to 5) for all of the questions.
The instrument did not include any conflicting, contradictory, or
reverse scoring questions to trap those respondents who consistently
provide biased responses to questionnaires irrespective ofwhat they really
believe to be the true answer (Paulhus, 1991; Smith, 2004).