diff --git a/wonkyconn/features/quality_control_connectivity.py b/wonkyconn/features/quality_control_connectivity.py
index d36c738..16a50ab 100644
--- a/wonkyconn/features/quality_control_connectivity.py
+++ b/wonkyconn/features/quality_control_connectivity.py
@@ -1,3 +1,4 @@
+from itertools import chain
 from typing import Iterable
 
 import numpy as np
@@ -33,29 +34,34 @@ def calculate_qcfc(
         pd.DataFrame: The QCFC values between connectivity matrices and the metric.
 
     """
-    metrics = pd.Series(
+    metrics = np.asarray(
         [
             connectivity_matrix.metadata.get(metric_key, np.nan)
             for connectivity_matrix in connectivity_matrices
         ]
     )
-    covariates = dmatrix("age + gender", data_frame)
+    covariates = np.asarray(dmatrix("age + gender", data_frame))
 
-    connectivity_array = np.concatenate(
-        [
-            connectivity_matrix.load()[:, :, np.newaxis]
-            for connectivity_matrix in tqdm(
-                connectivity_matrices, desc="Loading connectivity matrices", leave=False
-            )
-        ],
-        axis=2,
-    )
-    n, _, m = connectivity_array.shape
+    connectivity_arrays = [
+        connectivity_matrix.load()
+        for connectivity_matrix in tqdm(
+            connectivity_matrices, desc="Loading connectivity matrices", leave=False
+        )
+    ]
+
+    # Ensure that all arrays are square and have the same shape
+    (n,) = set(chain.from_iterable(a.shape for a in connectivity_arrays))
 
+    # Extract the lower triangles
     i, j = np.tril_indices(n, k=-1)
-    correlation = partial_correlation(
-        connectivity_array[i, j], metrics.to_numpy(), np.asarray(covariates)
+    connectivity_array = np.concatenate(
+        [a[i, j, np.newaxis] for a in connectivity_arrays],
+        axis=1,
     )
+
+    _, m = connectivity_array.shape
+    correlation = partial_correlation(connectivity_array, metrics, covariates)
+
     p_value = correlation_p_value(correlation, m)
 
     qcfc = pd.DataFrame(dict(i=i, j=j, correlation=correlation, p_value=p_value))