Quiz (Chapter 8, Second Part)


Significance Testing and the Hybrid Approach to Testing Ho

Introduction

  1. (T/F)  In all three versions of hypothesis testing (6-step, 7-step, 9-step), Ho is rejected if p< a.
  2. Significance testing is used _____ (more/less) often than hypothesis testing.

Significance Testing: It's Component Parts
  1. (T/F)  Significance testing involves fewer steps than even the 6-step version of hypothesis testing.
  2. Which of the following components of hypothesis testing is/are absent from significance testing?
    1. Ho
    2. Ha
    3. choice between a one/two-tailed test
    4. a
    5. collection of sample data
    6. computation of a test statistic
    7. determination of p
    8. a reject/fail-to-reject decision
  3. Can a researcher engage in statistical testing without initially deciding upon his/her study's statistical focus?
  4. (T/F) The Ho in significance testing takes the same form as the null hypothesis in hypothesis testing.
  5. In significance testing, the one-/two-tailed decision should be made ___ (before/after) the data are examined.
  6. (T/F)  In contrast to the p in hypothesis testing, the p in significance testing will be small if the data support Ho.
  7. In significance testing, is p normally reported precisely (e.g., p =.02) or in a "less-than" format (e.g., p<.05)?
  8. (T/F) In significance testing, the researcher does not end up rejecting or failing-to-reject his/her Ho.

Significance Testing Using "p-less-than" and "p-greater-than" Statements
  1. (T/F) When results are reported by a researcher who has performed significance testing, the symbol between p and a decimal number will always an equal sign (and never < or >).
  2. Can someone do significance testing if he/she doesn't have a computer and thus can't determine p precisely?

The Hybrid Approach: The Simple Idea
  1. (T/F)  True significance testing does not involve an a-level or a reject/fail-to-reject decision about Ho.
  2. (T/F)  True hypothesis testing does not permit the researcher to report "impressive" results (e.g., p<.001) if he/she has decided, in Step 3 of the process, to use the .05 level of significance.
  3. In the hybrid approach, what will the researcher report besides his/her reject/retain decision concerning Ho?
  4. If a researcher reports that his/her data revealed a "significant difference at p<.001" but doesn't say anything at all about using an a-level, what should you guess happened?
    1. He/she did significance testing
    2. He/she did hypothesis testing with a = .001
    3. He/she used the hybrid approach with a = .05

The Hybrid Approach: Terms for "Near Misses" and "Impressive" Findings
  1. If a researcher is using the .05 level of significance, which of these data-based p-levels might very well prompt him/her to report that the results "indicate a trend toward significance?"
    1. p = .003
    2. p = .048
    3. p = .054
    4. p = .377
    5. p = .983
  2. Answer #17 again 3 times, replacing the final phrase initially with the term "were marginally significant,"  next  with the term "approached significance" and finally with the term "were of borderline significance."
  3. If a researcher sets alpha at .05, which of these data-based p-levels might very well prompt him/her to report that the results were "highly significant?"  Answer again, now with "clearly significant" as the last 2 words.
    1. p = .003
    2. p = .048
    3. p = .054
    4. p = .377
    5. p = .983

The Hybrid Approach: Reporting Fail-to-Reject Decisions With "p-less-than" Statements
  1. (T/F) If you see "p < .30," this means that the null hypothesis was rejected at the .30 level of significance.
  2. (T/F)  A researcher might state that "p< .20" in order to communicate the finding that "p > .05."

The Hybrid Approach: The "One-Sided" Alternative
  1. (T/F) If a researcher uses the "one-sided" variation to the hybrid approach to testing null hypotheses, his/her  alternative hypothesis will automatically be directional in nature.
  2. In using the "one-sided" variation of the hybrid approach, which of these two things will researchers do?
    1. Use terms like "marginally significant" if p is slightly higher than a, but simply say "significant" if p < a
    2. Use terms like "highly significant" if p is way under a, but simply say "not significant" if p > a

The Hybrid Approach: The 7-Step, 8-Step, and 10-Step Options
  1. What gets "tacked on at the end" to the 6-, 7-, or 9-step versions of hypothesis testing to create the 7-, 8-, or 10-step versions of the hybrid approach to testing null hypotheses?

A Few Warnings
  1. Which approach--hypothesis testing, significance testing, or the hybrid approach to testing an Ho--has the ability to bring forth interesting results even if the null hypothesis is quite silly?
  2. (T/F) The practical significance of a study can be assessed by looking to see how small the p-value is.
  3. (T/F)  In the hybrid approach, what's considered a "near miss" or "highly significant" is highly subjective.
  4. (T/F) If p turns out equal to 0.0000, this means that the null hypothesis has been proven to be false.
  5. Who is ultimately in control of the level of significance associated with any use of the hypothesis testing procedure or the hybrid approach to testing null hypotheses:
    1. The research subjects
    2. The journal's editorial board
    3. The researcher
    4. The scientific community
    5. The recipient of the research report


Click here for answers.

 

Copyright © 2012

Schuyler W. Huck
All rights reserved.

| Book Info | Author Info |

Site URL: www.readingstats.com

Top | Site Map
Site Design: John W. Taylor V