Statistics Flashcards

1
Q

In the demographic characteristics, you found that the SRC group had a greater body weight and GLETQ score, exercised at a more intense workload, and maintained a higher steady-state HR than the HC group. How might these differences have influenced your results, and what steps did you take to control for these potential confounds?

A
  • The differences in demographic characteristics between the SRC and HC groups could potentially influence the results of the study. For example, the higher body weight and GLETQ score in the SRC group may indicate differences in fitness levels, which could affect the physiological responses to exercise (e.g., HR and MCAv) and potentially impact cognitive performance.
  • Additionally, the more intense workload and higher steady-state HR in the SRC group could lead to greater physiological arousal, which may influence oculomotor performance differently compared to the HC group.
  • However, despite differences in fitness levels and exercise responses (reflected by body weight, GLETQ score, workload, and HR) between the SRC and HC groups, the observed decrease in antisaccade RTs from pre- to post-exercise in both groups suggests a facilitatory effect of acute exercise on cognitive control processes, with similar benefits for individuals with and without SRC (i.e., no group by time interaction).

In the future, we would endeavour to match the groups on these characteristics during the recruitment process to minimize differences between the groups.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

For the HR and MCAv results, you computed difference scores (steady-state minus baseline) to uncover the nature of the interaction. Can you explain the rationale behind this approach and discuss any limitations of using difference scores?

A
  • We computed difference scores (steady-state minus baseline) for HR and MCAv to better understand the nature of the interaction effects found in the split-plot ANOVAs.
  • By calculating the change from baseline to steady-state, the difference scores provide a measure of the magnitude of change in HR and MCAv in response to exercise.
  • This approach can help to isolate the effects of the exercise intervention on these physiological variables and facilitate the interpretation of the interaction effects.
  • However, there are some limitations to using difference scores:
  • a) Difference scores can be less reliable than the original measures, as they compound the measurement error from both the baseline and steady-state assessments.
  • b) Difference scores may be affected by regression to the mean, where extreme values at baseline tend to be closer to the mean at follow-up, potentially leading to spurious conclusions about change over time.
  • c) Difference scores can be difficult to interpret when there are significant baseline differences between groups, as the magnitude of change may be influenced by the starting point.
    Despite these limitations, difference scores remain a commonly used approach in repeated measures designs to examine change over time and can provide valuable insights into the nature of interaction effects when used and interpreted appropriately.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

In the oculomotor performance results, you found that antisaccade RTs decreased from pre- to postexercise for both SRC and HC groups. How do you interpret this finding, and what are the potential implications for understanding the effects of exercise on cognitive function in individuals with and without SRC?

A
  • The finding that antisaccade RTs decreased from pre- to postexercise for both SRC and HC groups suggests that acute exercise may have a beneficial effect on cognitive control processes, specifically inhibitory control.
  • Antisaccade tasks require participants to inhibit the automatic response of looking towards a sudden-onset stimulus and instead generate a voluntary saccade in the opposite direction.
  • The reduction in antisaccade RTs following exercise indicates an improvement in the efficiency of executive function.
  • This finding has potential implications for understanding the effects of exercise on cognitive function in individuals with and without SRC.
  • It suggests that acute exercise may provide a transient boost to cognitive control abilities, which could be beneficial for both healthy individuals and those recovering from SRC.
  • For individuals with SRC, the improvement in antisaccade performance following exercise may indicate that supervised graded aerobic exercise could be a useful tool in rehabilitation to aid in the recovery of cognitive function.
  • However, it is important to note that this study only examined the acute effects of exercise, and further research is needed to understand the long-term implications and potential benefits of exercise for executive function in SRC recovery.
  • Furthermore, the lack of a significant interaction effect between group and time for antisaccade RTs suggests that the facilitatory effect of exercise on cognitive control may be similar for both healthy individuals and those with SRC, at least in the acute setting.
    This finding highlights the potential for exercise to be a valuable tool for improving cognitive function across different populations.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

You did not find a significant correlation between MCAv or HR difference scores and antisaccade RT difference scores. What are the possible reasons for this lack of correlation, and how does this impact the interpretation of your results?

A
  • The lack of a significant correlation between MCAv or HR difference scores and antisaccade RT difference scores can be attributed to several factors:
  • The physiological changes induced by exercise (i.e., increased MCAv and HR) may not be directly related to the cognitive processes involved in antisaccade performance.
  • This finding does not necessarily invalidate the main results of the study, as the split-plot ANOVAs demonstrated significant improvements in antisaccade RTs following exercise for both groups.
  • While exercise may lead to general improvements in inhibitory control, the specific mechanisms underlying the improvement in antisaccade RTs may be more complex and not captured by simple correlations with MCAv or HR changes.
  • The sample size of the study may have been insufficient to detect a significant correlation between these variables. Correlational analyses typically require larger sample sizes to have adequate statistical power to detect small to moderate effect sizes.
  • Individual differences in fitness levels, exercise responses, and cognitive abilities may have introduced variability into the data, making it more difficult to detect significant correlations at the group level.
  • Further research is needed to investigate the potential mediators of the relationship between exercise and cognitive function, as well as to explore the individual differences that may influence these associations. This may involve:
  • Using more comprehensive physiological and neuroimaging measures
  • Assessing cognitive function during exercise
  • Recruiting larger sample sizes to increase statistical power for correlational analyses
  • These steps will help clarify the mechanisms through which exercise affects cognitive performance and better account for individual variability.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

For the perceived exertion and symptomology results, you found that SCAT-5 symptom frequency and severity did not differ between pre-and postexercise oculomotor assessments. How do you reconcile this finding with the improvements observed in antisaccade RTs, and what are the potential clinical implications?

A
  • The finding that SCAT-5 symptom frequency and severity did not differ between pre- and post-exercise oculomotor assessments suggests that the acute exercise intervention did not exacerbate concussion-related symptoms in the SRC group. This has important clinical implications:
  • These findings suggest that carefully prescribed and monitored exercise can be a valuable tool in the rehabilitation of individuals with SRC, even if they have ongoing symptoms.
  • The lack of significant changes in SCAT-5 symptom frequency and severity supports the notion that graded exercise testing and progression can be safely implemented in this population, provided that appropriate guidelines and monitoring procedures are followed.
  • The observed improvements in antisaccade RTs following exercise might seem at odds with the stable SCAT-5 scores. One possibility is that the SCAT-5 may not be sensitive enough to detect subtle changes in symptom frequency and severity, particularly in cognitive domains related to antisaccade performance.
  • The SCAT-5 is a relatively coarse measure and may not capture nuanced changes in cognitive or physiological functioning that could be associated with improvements in antisaccade RTs.
  • Future studies should consider using more sensitive measures to detect subtle changes in cognitive or physiological symptoms post-exercise.
  • Conducting cluster-specific analyses of SCAT-5 symptomology could provide more detailed insights into how different symptom domains are affected by exercise.
    Broader use of comprehensive neuropsychological and physiological assessments can help better understand the relationship between exercise, symptom burden, and cognitive performance in SRC populations.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Can you explain the significance of the main effect for time for RPE in the split-plot ANOVA, particularly the linear increase in values observed during the exercise intervention?

A

The significant main effect of time in the split-plot ANOVA, with values increasing linearly during the exercise intervention, indicates that perceived exertion progressively increased as the exercise session went on for both groups. This is an expected finding, as physical effort and fatigue build up over the course of an exercise session.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What does the absence of a reliable main effect for group suggest about the differences in perceived exertion between the SRC and HC groups during the exercise intervention?

A
  • The lack of a significant main effect for group suggests that overall, the SRC and HC groups did not differ in their perceived exertion levels during the exercise intervention. This implies that the concussion history of the SRC group did not significantly influence their perception of exercise intensity compared to healthy controls.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

The group by time interaction was not significant. What are the implications of this result for the relationship between group membership and perceived exertion over time?

A
  • The non-significant group by time interaction indicates that the pattern of change in perceived exertion over time was similar for both the SRC and HC groups. In other words, the progression of perceived exertion during the exercise session did not differ based on group membership (SRC vs HC).
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Discuss the results of the SCAT-5 total symptom frequency and severity evaluations. Specifically, what do the non-significant differences between pre-and postexercise oculomotor assessments indicate?

A

The non-significant differences in SCAT-5 symptom frequency and severity between pre- and post-exercise oculomotor assessments suggest that the exercise intervention did not lead to a significant exacerbation of concussion-related symptoms in the SRC group immediately after the session.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

The symptom frequency and severity were significantly less at the 24-hour follow-up compared to the pre-BCBT assessment. What might account for this reduction in symptoms?

A
  • The significant reduction in symptom frequency and severity at the 24-hour follow-up compared to the pre-BCBT assessment could be attributed to the natural recovery process after concussion. While it is possible that the controlled exercise intervention did not hinder recovery and may have even facilitated it, this interpretation should be made with caution.
    Without a control condition (i.e., SRC group that did not engage in exercise), it is difficult to definitively attribute the symptom reduction to the exercise intervention alone. Factors such as time since injury and individual differences in recovery trajectories should also be considered.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Why did you choose to conduct power polynomials (i.e., trend analysis)

A
  • In this study, we conducted power polynomials (i.e., trend analysis) to gain a detailed understanding of the pattern of change in perceived exertion (RPE scores) over time during the exercise intervention. There are several reasons for this choice:
  • Complementing the ANOVA Results:
  • While the split-plot ANOVA revealed a significant main effect of time, indicating that RPE scores changed significantly throughout the exercise session, it did not provide information about the specific pattern of this change. By conducting trend analysis using power polynomials, we were able to complement the ANOVA results and gain a more detailed understanding of how perceived exertion changed over time.
  • Examining Non-linear Trends:
  • Power polynomials allow us to examine non-linear trends in the data. In this study, the significant quadratic polynomial trend suggests that the change in RPE scores over time was not entirely linear. The quadratic trend captures the initial increase in perceived exertion during the exercise phase, followed by a decrease during the cool-down period. This non-linear pattern provides a more nuanced understanding of how perceived exertion changes over the course of the exercise intervention.
  • Meaningful Interpretation:
  • Trend analysis using power polynomials allows for a more meaningful interpretation of the data by identifying specific patterns of change. In this study, the significant linear and quadratic trends provide insights into the progression of perceived exertion throughout the exercise session. The linear trend indicates a general increase in RPE scores over time, while the quadratic trend captures the more specific pattern of an initial increase followed by a decrease during the cool-down phase.
    By using power polynomials, we were able to uncover these detailed patterns, enhancing our understanding of the dynamics of perceived exertion during the exercise intervention.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

How does the quadratic polynomial result aid in understanding the pattern of perceived exertion during the exercise intervention?

A
  • The significant quadratic polynomial result aids in understanding the pattern of perceived exertion during the exercise intervention by revealing a curved trajectory.
  • The quadratic trend shows an initial increase in perceived exertion during the main exercise phase, followed by a decrease during the cool-down period.
  • This pattern is consistent with the expected progression of perceived exertion during a typical exercise session, where effort and intensity build up during the main exercise phase and then gradually decrease during the cool-down.
    The quadratic polynomial result captures this non-linear pattern and provides a more nuanced understanding of how perceived exertion changes over the course of the exercise intervention.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Analyze the Demographic Table and discuss between-group differences

A
  • Age and Sex:
    ○ The average age of participants in both groups is similar, with SRC at 20.5 years and HC at 20.9 years.
    ○ Both groups have the same number of females, 5 out of 16 participants (31.3%).
  • BMI:
    ○ No significant BMI difference: SRC (26.5) vs HC (22.5).
  • Height and Weight:
    ○ No significant height difference SRC (177.2 cm) vs HC (174.6 cm)
    ○ Significant Weight difference (84.1 kg) vs. (68.9 kg).
  • Heart Rate (HR):
    ○ Baseline HR is similar between the groups, with SRC at (73.1 bpm) and HC at (72.3 bpm)
    ○ The SRC group has a higher steady-state HR (127.2 bpm) compared to the HC group (114.6 bpm).
    ○ Post-exercise HR is non-significantly lower in the SRC group (83.1 bpm) compared to the HC group (90.9 bpm).*

Blood Pressure (SBP/DBP):
○ Resting SBP/DBP values are similar between the groups, with SRC at (117.3/76.5 mmHg) and HC at (115.8/74.3 mmHg).
○ Post-exercise SBP/DBP values are also similar, with SRC at (121.4/77.4) mmHg and HC at (121.9/74.6 mmHg).

  • Heart Rate Threshold (HRt):
    ○ Both groups have the same HRt at (120.0 bpm).
  • Workload:
    ○ The SRC group has a significantly higher average workload (93.3 W) compared to the HC group (69.5 W).
  • GLETQ Total Score:
    ○ The SRC group has a significantly higher GLETQ total score (84.4) compared to the HC group (65.1), indicating more frequent physical activity.
  • VAS and RPE Scores:
    ○ VAS Score Visit 1: SRC 1.6 mean score
    ○ RPE Score Visit 1: SRC 11.1 mean score “Conversation is easy and I can exercise like this for a while”
    ○ VAS Score Visit 2: SRC group has a mean score of (0.5) between “No symptoms” and “Feel some symptoms”, while HC group had no symptoms.
    ○ RPE Score Visit 2: Both groups have similar scores, with SRC at a mean of (9.2) and HC at (9.3). This corresponds to “Very Light Exertion”
  • Concussion History:
    ○ The SRC group has an average of 2 previous concussions.
  • Days Since Injury:
    ○ Days Since Injury to Visit 1: 8.3 days.
    ○ Days Since Injury to Visit 2: 10.1 days.
    Days Since Visit 1 to Visit 2: 1.8 days.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Justify your decision to not perform a post-hoc test for the ANTI RT difference scores between groups.

A

The decision to not perform a post-hoc test for the ANTI RT difference scores between groups was reasonable to avoid increasing the risk of a Type I error, as we had a clear a priori hypothesis and wanted to minimize the number of statistical comparisons. This approach helps maintain the integrity of the hypothesis testing and reduces the likelihood of inauthentic findings.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Why are effect size measures, such as partial eta-squared (η²p) and Cohen’s d, important for interpreting the study findings?

A
  • Effect size metrics, such as partial eta-squared (η²p) and Cohen’s d, are essential for providing practical significance to the statistical results. While p-values indicate the likelihood of the data under the null hypothesis, effect sizes quantify the magnitude of the observed effects, which is crucial for understanding the real-world importance and potential applications of the findings.
  • Partial eta-squared allows assessment of the relative strength of different factors, beyond just their statistical significance. Large η²p values indicate the factors explained substantial variability, while smaller values suggest more modest group differences.

Cohen’s d provides a standardized metric for evaluating the practical significance, enabling comparisons to existing literature and informing future study design. The large d-values underscore the meaningful cognitive changes induced by the concussive injury and exercise.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Explain the differences in degrees of freedom between group and within group and what would they be in your sample

A
  • df = 15
    ○ Applies to the within-group df for a single group (SRC or HC) with n = 16 participants
    ○ Calculated as: df = n - 1 = 16 - 1 = 15
    ○ Used for:
    § Analyzing variability within each individual group (SRC or HC)
    • df = 30
      ○ Applies to:
      § The combined within-group df for the two groups (SRC and HC) with total N = 32
      § The df for an independent samples t-test comparing the two groups
      ○ Calculated as: df = N - k = 32 - 2 = 30
      ○ Used for:
      § Analyzing variability across the combined SRC and HC groups
      § Conducting the independent samples t-tests comparing the SRC and HC groups
    • df = 31
      ○ Applies to:
      § Single-sample t-tests for the total sample of N = 32
      § Correlations involving the entire sample of N = 32
      ○ Calculated as: df = N - 1 = 32 - 1 = 31
      ○ Used for:
      § Conducting single-sample t-tests on the overall sample
      § Calculating the Pearson’s r correlations involving the pooled SRC and HC data
17
Q

Explain the differences in degrees of freedom between group and within group and what would they be in your sample

A
  • df = 15
    ○ Applies to the within-group df for a single group (SRC or HC) with n = 16 participants
    ○ Calculated as: df = n - 1 = 16 - 1 = 15
    ○ Used for:
    Analyzing variability within each individual group (SRC or HC)
  • df = 30
    ○ Applies to:
    The combined within-group df for the two groups (SRC and HC) with total N = 32
    The df for an independent samples t-test comparing the two groups
    ○ Calculated as: df = N - k = 32 - 2 = 30
    ○ Used for:
    Analyzing variability across the combined SRC and HC groups
    Conducting the independent samples t-tests comparing the SRC and HC groups
  • df = 31
    ○ Applies to:
    Single-sample t-tests for the total sample of N = 32
    Correlations involving the entire sample of N = 32
    ○ Calculated as: df = N - 1 = 32 - 1 = 31
    ○ Used for:
    Conducting single-sample t-tests on the overall sample
    Calculating the Pearson’s r correlations involving the pooled SRC and HC data
18
Q

For a post-hoc test, you ran pairwise t-tests. Discuss the limitations of this approach

A
  • Conduct the ANOVA first.
  • If the ANOVA is significant, perform pairwise t-tests between all groups.
  • Advantages:
  • Conceptually simple and easy to understand.
  • Provides specific p-values for each comparison.
  • Limitations and Considerations:
  • Increased Type I Error Rate:
    ○ Running multiple t-tests increases the chance of finding a significant result by chance (Type I error).
    ○ The more comparisons you make, the higher this risk becomes.
  • No Built-in Correction:
    ○ Unlike many dedicated post hoc tests, simple t-tests don’t automatically adjust for multiple comparisons.
  • Need for Manual Correction:
    ○ To control the family-wise error rate, you need to manually apply a correction method (e.g., Bonferroni, Holm-Bonferroni).
  • Less Power:
    ○ After applying corrections, this method can be less powerful than some dedicated post hoc tests.
  • Doesn’t Use Pooled Variance:
    ○ ANOVA uses a pooled estimate of variance, which can be more precise. Individual t-tests don’t benefit from this.
  • Not Recommended in Many Fields:
    ○ Many disciplines and journals prefer established post hoc procedures over multiple t-tests.