One-tail test for research

#1
Hello,
I am new to research metholologies and statistical testing. Currently, I am involved in the data analysis section of research. My instructor wants me to discuss the application of a one-tail test to my research. I am studying the effect of technology training on teachers' computer self-efficacy and technology practices. The hypothesis is H01: There is no statistically significant relationship between technology training and teachers’ perceptions of computer self-efficacy (PCSE), current instructional practices (CIP), level of technology implementation (LoTi), and personal computer use (PCU).
The alternative hypothesis is, Ha1: There is a positive statistically significant relationship between technology training and teachers’ perceptions of computer self-efficacy (PCSE), current instructional practices (CIP), level of technology implementation (LoTi), and personal computer use (PCU). With a limited understanding of statistics, I need to understand the one-tail test and its application to my study. Any help will be much appreciated.
 

CB

Super Moderator
#2
Hi there, welcome.

First comment: the phrase "statistically significant" shouldn't really appear in your framing of H0 and H1. H0 and H1 are specific statements about a population. The significance of relationships found in your sample may allow you to make inferences about the population of interest.

So for a one-tailed test your hypotheses might be that in the population of interest:

H0: The correlation between technology training and teachers’ perceptions of computer self-efficacy (PCSE) is zero
H1: The correlation between technology training and teachers’ perceptions of computer self-efficacy (PCSE) is greater than zero

You may need to specify H0 and H1 for each relationship that you're going to look at. Keep a couple of things in mind:

1. A null hypothesis of an exactly zero effect or relationship is rarely plausible in the social sciences, so rejecting this hypothesis isn't particularly groundbreaking. The magnitude of the relationship is probably of more interest: you might want to supplement significance testing with confidence intervals, for example.
2. A one-tailed test means you're disregarding the possibility of a negative relationship between the variables. Are you confident that you can rule out this possibility?